Skip to main content

Full text of "The Philippine journal of science"

See other formats

HH i y yij i | i li pi |ipi i yi |il i| i 

Ri:<.i:ivEi) IN i'^xciiANcsi: 


Philippine Island Bur. 


The Philippine 
Journal of Science 

Volume 56 







EuLOGio Rodriguez, A.B,, Secretary 
Jorge B, Vargas, A.B., LL.B., Under Secretary 


Published by the Bureau of Science, Department of Agriculture 

and Commerce 

[Entered at the Post Office at Manila, P. I., as second-class matter.] 

A. S. ARGt^ELLES, B.S., Editor 

Leopoldo a. Faustino, E.M., Ph.D., Associate Editor 

Richard C. McGregor, A.B., Managing Editor 

contributing editors 


Manuel L. Roxas, Ph.D.; A. P. West, Ph.D.; T. Dar Juan, Phar.D. 

F. D. Reyes, B.S.; F. T. Adriano, Ph.D.; R. H. Aguilar, Ch.E. 

A. J. Hermano, ScD.; J. C. Espinosa, B.S. in Ch.E. 

Maria Y. Orosa, Ph.C, M.S. 

V. Elicano, B.S.; Q. A. Abadilla, E.M.; Jose M. Feliciano, Ph.D. 

Experimental Medicine 
Onopre Garcia, M.D.; H. W. Wade, M.D.; Artubo Garcia, M.D. 

Daniel de la Paz, M.D.; Cristobal Manalang, M.D. 
Victor Bubncamino, D.V.M.; Teodulo Topacio, D.V.M., D.Sc. 

Clinical Medicine 

LiBOBio Gomez, M.D., Ph.D.; F. Calderon, L.M. 

Jacobo Fajardo, M.D.; Jos^ Albert, M.D.; H. Lara, M.D. 

Josig Rodriguez, M.D.; Carmelo Reyes, M.D. 


William H. Brown, Ph.D.; J. K. Santos, Ph.D.; P. L. Sherman, Ph.D. 

Leon Ma. Guerrero, Phar.D.; A. F. Fischer, C.E., M.F. 

Eduardo Quisumbing, Ph.D.; T. G. Fajardo, Ph.D. 

Joaquin Maranon, Sc.D; Rafael B. Espino, Ph.D. 

Nicanor G. Teodoro, Ph.D.; Vicente G. Aldaba, Ph.D. 

Feliciano M. Clara, Ph.D.; Juan P. Torres, Ph.D.; Luis P. Reyes, B.S.F. 


Marcos A. Tubangui, D.V.M.; Heraclio R. Montalban, M.A. 

Leopoldo B. Uichanco, Sc.D.; Gonzalo Merino, Ph.D. 

Faustino Q. Otanes, M.S.; Hilario A. Roxas, Ph.D. 

Manuel D. Sumulong, M.S., D.V.M.; Lope M. Yutuc, D.V.M. 

Deogracias Villadolid, Ph.D.; Canuto G. Manuel, Sc.D. 

H. O. Beyer, M.A,; Otto Johns Scheerer, M.A.; E. E. Schneider, B.L. 


No. 1, January, 1935 

tissued March 29, 1935.] 


Africa, Candido M., and E. Y. Garcia. The occurrence of Bertiella 

in man, monkey, and dog in the Philippines 1 

Four plates. 

Tubangui, Marcos A. Additional notes on Philippine Acanthoce- 

phala 13 

Two plates. 

SUGINO, KouHEi. Studies on the diastolic blood pressure in beriberi 
cases admitted to the Philippine General Hospital during the 
years 1932-1934 21 

Blanco, Guillermo J. The Atyida^ of the Philippine Islands 29 

Three plates. 

Blanco, Guillermo J. The development of the homocercal caudal of 

the blue perch, Taeniotoca lateralis Agassiz 41 

One plate. 

Stevens, F. L., and E. F. Roldan. Philippine Meliolin^e 47 

Three text figures. 

Galang, Ricardo E. Ethnographic study of the Yogads of Isabela 81 

Thfee plates. 

Manuel, Canuto G. New birds from northern Luzon, Philippine Is- 
lands ^^ 

One plate. 

No, 2, February, 1935 

[Issued May 9, 1935.] 

Copeland, E. B. New or interesting Philippine Ferns, VIII 97 

Fourteen plates. 

Serrano, F. B. Control of pineapple mealy-bug wilt Ill 

Two plates. 

TOKUNAGA, Masaaki. A morphological study of a nymphomyiid fly- 127 

Five plates and six text figures. 

Martin, Claro, and Heraclio R. Montalban. Philippine Paraper- 

cid« 215 

Three plates. 

No. 3, March, 1935 

[Issued June 11, 1935.] 

Maranon, Joaquin, Amando Perez, and Paul F. Russell. Phil- 
ippine Totaquma 

Two text figrures. 

Russell, Paul F., and Francisco E. Baisas. The technic of hand- 
ing mosquitoes 

Eight plates and twelve text figures. 


291468 12 

iv Contents 

Garcia, Eusebio Y. Effects of chlorinated lime in lethal concentra- 
tions on Entamoeba histolytica cysts 295 

One plate. 

Merrill, Elmer D., and Eduardo Quisumbing. Styrax in the Phil- 
ippines 313 

One plate. 

Umali, Agustin F. Little-known fishes from the Philippines 319 

One plate. 

De Laubenfels, M. W. A collection of sponges from the Marine 

Station at Puerto Galera (Mindoro) Philippine Islands 327 

One plate. 

Alexander, Charles P. New or little-known Tipulidae from eastern 

Asia (Diptera), XXIII 339 

Three plates. 

Baumgartner, W. J., and Marcellus T. Surla. The spermatid 
transformation in Amblycorypha oblongifolia (de Geer), a tet- 
tigoniid 373 

Two plates. 

Schedl, Karl E. Fauna philippinensis (Platypodidae et ScolytidsB), 

III 395 

No. 4, April, 1935 

[Issued June 29, 1985.] 

JULIANO, Jose B. Anatomy and morphology of the buiiga, Aeginetia 

indica Linnaeus 405 

Ten plates. 

Ames, Oakes, and Eduardo Quisumbing. New or noteworthy Philip- 
pine orchids, V 453 

Ten plates. 

COPELAND, E. B. Additional ferns of Kinabalu 471 

Ten plates. 

Baisas, F. E. Notes on Philippine mosquitoes, I: The Armigeres 

group - 485 

Four plat^ and three text figures. 

Takahashi, Ryoichi. Additions to the aphid fauna of Formosa 

(Hemiptera) , III 499 

Three text figures. 

Voss, Eduard. Einige unbeschriebene Curculioniden aus dem indo- 
malayischen Archipel. 58. Beitrag zur Kenntnis der Curculio- 
niden 509 

One text figure. 

ALEXANDi3R, CHARLES P. New or little-known Tipulidse from eastern 

Asia (Diptera), XXIV 525 

Three plates. 

Erratum 563 

Index 565 

The Philippine 
Journal of Science 

Vol. 56 JANUARY, 1935 No. 1 


By Candido M. Africa and E. Y. Garcia 

Of the School of Hygiene and Public Health, University of the Philippinei 



Early in 1934 we received from Doctor Gustilo for identifica- 
tion a tapeworm, which he recovered from his eight-year-old 
daughter, who had always resided in Sara, Iloilo Province, 
Panay. The tapeworm was a specimen of Bertiella, The gen- 
eral appearance of this tapeworm reminded us of a similar 
specimen that we had recovered at autopsy from the small intes- 
tine of a Manila dog five months previously. This specimen 
proved to be a Bertiella also, although it differs somewhat 
morphologically from our human material. While study of 
these two specimens was in progress, we received from Dr. 
Marcos Tubangui, of the Bureau of Science, Manila, three com- 
plete strobilse of an apparently similar tapeworm, which he 
recovered at autopsy from the small intestine of a Philippine 
monkey (Macacus cynomolgus) in Rizal Province, near Manila, 
October 12, 1932. A study of this monkey material convinced 
us that, except for some insignificant difference, it is identical 
with our human material. 

The interest that may be attached to this report is threefold : 
(a) a new area in the nosogeography of Bertiella is established ; 
(6) a new case of Bertiella infestation in man is recorded; (c) 
the occurrence of Bertiella in the dog adds a new animal to the 
list of hosts for this genus. For convenience we shall call our 

289096 1 

2 The Philippine Journal of Science i^ss 

three lots of material by the names of their respective hosts in 
referring to them in this paper, 


The human tapeworm, which was preserved in alcohol, ap- 
peared to be a complete strobila with the scolex missing. The 
entire specimen measures 160 millimeters in length.^ The tape- 
worm broadens and thickens very gradually from the neck re- 
gion posteriorly, being widest and thickest at about the middle 
of the specimen. Its broadest segments measure 11 millimeters 
in width, 0.2 millimeter in length, and 3 millimeters in thickness. 
The hindermost segments measure 7 millimeters in width, 0.45 
millimeter in length, and 1.2 millimeters in thickness. The ripe 
segments are apparently shed not singly but in groups, since 
previous to the expulsion of this strobila a fragment of the tape- 
worm consisting of nine ripe segments was passed by the child. 

Segments taken from various levels of the specimen were 
stained after the method of Ristroph (1934). Individual seg- 
ments were also isolated, stained, and mounted in toto lying flat 
on their anterior or posterior border. This technic affords an 
unobstructed view of the internal structure, especially the re- 
productive glands and tubes, as it does away with the necessity 
of looking through the calcareous deposits and musculature, 
which obstruct the view when the segment is examined lying 
on its dorsal or ventral surface. Serial sections, longitudinal 
and transverse, of both mature and ripe segments stained after 
the above method were also prepared for study. 

No description of the head can be given as this was missing 
in the specimen. The segments are broader than long; each 
segment contains a single set of reproductive organs ; the genital 
atria alternate irregularly and are situated within the anterior 
half of the lateral margin of the segments. The dorsal excretory 
canal is smaller than the ventral excretory canal; the genital 
canals are dorsal to both the dorsal and ventral longitudinal 

Female reproductive organs. — The ovary is saddlelike in 
shape with its back directed ventrally when viewed laterally. 
Measuring about 1.8 millimeters from tip to tip in a transverse 
direction, it occupies about one-fourth of the whole width of the 

* According to Dr. Gustilo the tapeworm when alive was capable of ex- 
tending its length to as much as 16 inches, or approximately two and one- 
half times that of the preserved specimen. 

56,1 Africa and Garcia: Bertiella 3 

mature segment, and is located well within the poral half of the 
proglottid. It consists of two large lobes connected with each 
other by a glandular bridge which communicates with a com- 
mon stem. Each lobe consists of many long, fingerlike, tubular- 
sacular cytogenic glands with enlarged ends apparently origin- 
ating from a common center. The vitellarium is also bilobed, 
and to some extent a copy of the ovary in general form and 
the relation of the lobes to each other. The lobes lie dorsally 
and behind the ovary, and being composed of numerous oval 
lobules with a tendency to form clublike ends radiating from 
a common center, may be mistaken for parts of the ovary. The 
shell gland is oval, situated dorsally and slightly anteriorly to 
the vitellarium. From this organ there arises a coiled tube, 
which empties into the receptaculum seminis. The recepta- 
culum seminis, which is a thin-walled, oval vesicle measuring 
around 450 by 310 [x, lies dorsally and porally to the ovary, and 
anteroporally to the vitellarium, its ventral half being obscured 
by the latter organs when viewed from behind. The almost 
straight and slender vagina arises from it abruptly, like the 
stem of a pipe. At its point of origin may be seen clumps of 
deeply staining cells. The ovarian portion of the vagina, which 
dilates slightly after emergence from the receptaculum seminis, 
is about 900 \l long and 35 [a wide, while its muscular and glan- 
dular poral segment is about 400 [i. long and 130 pi. wide. In serial 
sections the poral portion of the vagina appears surrounded by 
a layer of deeply staining glandular cells, which seem to radiate 
from the lumen. The vagina, whose muscular lumen is plugged 
with a yellow pigmentlike material, lies alongside the cirrus in 
a common stroma. It opens into the common genital atrium 
behind and ventrally to the cirral opening. The uterus is in 
the form of a simple wavy tube across the segment with its 
trough and crest directed dorsoventrally. 

In the ripe segment the medullary field is almost completely 
filled by the uterus, which is stretched transversely up to the 
lateral excretory vessels on either side in a corkscrew fashion. 
In serial sections it appears as a large tube with outpocketings 
packed completely with eggs that impart to it the appearance 
of a polygonal mosaic. The poral portion of the vagina persists 
in the ripe segment. 

The eggs are roughly globular bodies with the following meas- 
urements: Outer eggshell, 55 to 73 |x in diameter; inner eggshell, 
40 to 50 [A ; embryo, 13 by 25 jx; booklets, 9 to 10 {x. Freshly ob- 

4 The Philippine Journal of Science 1935 

tained eggs from the uterus have their outer shell more or 
less crinkled; the embryo and the inner shell are obscured by 
coarse, ref ractile granules of varying size. After clearing, how- 
ever, with Puri's fluid, the hexacanth embryo with its thick 
unstriated inner shell appears in full view. The egg thus 
cleared reminds one of the egg of Hymenolepsis nana. The 
pyrif orm apparatus is invariably present. It has the following 
average measurements: Length from base, 8 to 10 [x; width 
of base, 10 to 12 p.; distance of tip from outer eggshell, 15 

to 18 lA. 

Male reproductive organs. — The vas deferens runs towards 
the poral margin from the dorsal vicinity of the receptaculum 
seminis in three or four powerful sacculated loops before entering 
the thin-walled cirrus pouch, which lies dorsally and anterior to 
the poral portion of the vagina on the same plane as the ex- 
cretory vessels. The vas deferens dilates immediately after 
entering the cirrus sac into a more or less globular, thin-walled 
seminal vesicle, which appears glandular at its junction with 
the straight, uncoiled, fairly muscular but slender cirrus that 
runs alongside of, and in close contact with, the vagina to the 
common genital atrium close to the opening of the vagina, some- 
what dorsad and anteriad to the latter. The cirrus is unarmed 
and apparently unprotrusible. The testes are round or oval 
bodies, 78 to 80 ^ by 60 to 70 [x, numbering from 200 to 250, 
disposed in one to eight layers, confined in the anterodorsal 
field of the segment. No testes are found on the ventral side 
of the vagina. 


Our monkey Bertiella, preserved in 5 per cent formalin, is 
represented by three apparently complete strobilse in each 
of which the scolex is present. It has the following measure- 
ments : Total length, 158 millimeters ; greatest width, 14 ; great- 
est thickness, 4. Flattened and stained scolex measures 630 
by 510 [x. The suckers have an average diameter of 255 pi. 
No pigmentation is noticed anywhere in the head. On com- 
paring this material with our human Bertiella we noted that 
one is practically the copy of the other, except that the eggs 
of the human specimen are larger than those of the monkey 
specimen, a difference which we attribute to the different pre- 
servatives used. 

56,1 Africa and Garcia: Bertiella 5 


Our material preserved in 5 per cent formalin consists of 
two specimens, a smaller, apparently immature worm about 50 
millimeters long, and a larger, mature, and complete strobila, 
which has a total length of 190 millimeters. The short neck is 
immediately followed by the segments, which broaden and thick- 
en very gradually posteriad. The unflattened scolex is sub- 
globular and measures 900 by 650 p.. When flattened and stained 
it measures 820 by 560 [x. The proglottids are broader than 
long, and each contains a single set of reproductive organs ; the 
genital pores alternate irregularly with a tendency to be uni- 
lateral. The mature segment is 0.7 millimeter long, 7.0 milli- 
meters wide, and 1 millimeter thick. The dorsal excretory 
canal is smaller than the ventral excretory canal, and the genital 
tubes are both dorsal to both dorsal and ventral excretory canals. 

Female reproductive organs. — The bilobed ovary, which meas- 
ures 1.4 millimeters transversely, is crescent-shaped and located 
well within the poral half of the segment, occupying one-fifth of 
the width and almost the whole length and thickness of the pro- 
glottid. In transverse sections the two lobes appear to be joined 
together by a glandular bridge, which arises from a common 
stem. Each ovarian lobe hugs the corresponding lobe of the 
yolk glands which lie behind but in the same sagittal plane as 
the ovary. Sandwiched between the two lobes of the yolk 
glands lies the roundish shell gland, which sends out a short 
coiled tube that empties into the pyrif orm receptaculum seminis 
(300 by 135 [a), the latter being obscured by the poral lobe of 
the ovary. The vagina arises abruptly from the poral end of 
the receptaculum seminis and runs porally almost in a straight 
line until it reaches the plane of the excretory vessels, when it 
enlarges and becomes glandular and muscular to form its poral 
portion. This portion of the vagina runs alongside of the cirrus 
pouch and in close contact with it. 

Male reproductive organs.— The testes are round or ovoid 
bodies, 70 by 60 (x, numbering between 75 and 100, confined in 
the anterodorsal field of the segment. Although a number of 
the testes are found on the poral side of the ovary, none is 
found ventrally to the vagina. The vas deferens arises from 
the vicinity of the poral lobe of the ovary, and after running 
poralwards in mighty convolutions, enters the pyriform cirrus 

6 The Philippine Journal of Science 1935 

pouch (400 by 90 [J.) , which lies dorsally and behind the vagina. 
The seminal vesicle is a long cylindrical organ, which lacks the 
glandular elements found in the poral segment of the same 
organ in our human and monkey specimens. The apparently 
unprotrusible cirrus is long, slender, and straight. It empties 
into a short cloacal canal in common with the vagina before 
reaching the common genital atrium. 

The description of the ripe segment of our human and monkey 
material applies to the dog specimen. 

Although there are marked discrepancies between our dog 
material on the one hand and our human and monkey specimens 
on the other, we believe that the variations are not of specific 

According to Cram (1928) the genus Bertiella is composed 
at present of three species occurring in birds and nineteen spe- 
cies in mammals. The mammalian hosts are represented by 
rodents, marsupials, lemurs, and primates. However, a gen- 
eral review of the literature on the subject strongly indicates 
that many of these species are invalid or are subspecies of not 
more than two species. Thus, Baer (1927), quoted by Adams 
and Webbs (1933), reduces the recorded species of Bertiella to 
two — namely, J5. studeri of the Old World and B. mucronata of 
the New — although the probability that B. mucronata is identical 
with J?, studeri is pointed out by Cameron (1929). Southwell 
(1930) regards Bertiella satyri (Blanchard, 1891) Stiles and 
Hassal, 1902, Bertiella conferta (Meyner, 1895), Bertiella po- 
lyorchis Linstow, 1905, and Bertiella cercopitheci Beddard, 
1911, as synonyms of Bertiella studeri (Blanchard, 1891) Stiles 
and Hassal, 1902. Joyeux and Baer (1929) also maintain the 
synonymy of B. satyri and B. studeri. Meggit (1927) considers 
the differences between B, cercopitheci, B. mucronata, B. confer- 
ta, B, polyorchis, and B. studeri to be nonspecific and such as 
might be produced by variations in the methods of fixation or 
degree of muscular contraction of the worms. It seems that 
Bertiella fallax alone, described by Meggit (1927) from Cebus 
capuchinus in Egypt, presents undoubted differential specific 
characters from the type species, B. studeri. 

It appears, therefore, that the human infestation by this genus 
so far recorded is represented by only one species ; namely, Ber- 
tieUa studeri (Blanchard, 1891) Stiles and Hassal, 1902 (ssna- 
onym, B. satyri). 

Nine cases have been recorded so far of human infestations 
with B. studeri; namely, Bertiella satyri (Blanchard, 1891) re* 

6e,i Africa and Garcia: BertieUa 7 

covered first from the orang-outang {Simia satyri) in Borneo, 
and later reported twice in man, first from a child in Mauritius 
(Blanchard, 1913) and then from a Bengali child (Chandler, 
1925) ; the third case (B. satyri) was recorded from India 
(Mukerji, 1927) as occurring in a Hindu subject, although no 
account of the case was given; the fourth case appears to be 
B. miicronata (Meyner, 1895) Beddard, 1911, originally de- 
scribed from a Paraguayan black howler {Alloutta caraya) and 
reported by Cram (1928) as occurring in a young Spaniard 
in Cuba as well as in three young chimpanzees ; the fifth record 
(B. studeri) was from St. Kitts (Cameron, 1928) ; the sixth 
(B. studeri) again from India (Maplestone, 1930), although in 
this case there seems to be some confusion about the uterus 
which in the author's text figure is represented as directly con- 
nected with, or rather emptying into, the deferent duct; the 
seventh case (B. stvderi) was reported from Sumatra by^ Joyeux 
and DoUfus in 1931 (cited by Adams and Webbs, 1933) ; the 
eight and ninth cases {B, studeri) again in Mauritius (Adams 
and Webbs, 1933), the first from an 8.5-year-old Creole child, 
and the second from a 4.5-year-old girl of Indian parentage 
borne in Mauritius. Our human case of BertieUa infestation 
represents the tenth and newest record of this genus in man 
and promises to involve a species different from the type species, 
B, studeri. 

On comparing our human and monkey specimens with the 
type species, B. studeri (synonym, B. satyri) as described in 
Baer's account (Cameron, 1929), we found that our material 
differs particularly from the latter in the following respects: 
(a) The deferent canal of our specimens makes powerful sac- 
culated coils before entering the cirrus sac, whereas in B. studeri 
the vas deferens is almost a straight tube before its entrance 
into the same organ; (&) the number of testes in our specimens 
is from 200 to 250, disposed in from two to eight layers— in B. 
studeri it is from 150 to 300, disposed in one or two layers; {e) 
the eggs of our specimens measure 55 to 73 [x, with a mean of 
±L 64.53 ti., and the embryos +16.6 pi, whereas in B. studeri they 
measure from 45 to 60 i^; the embryos, 10 to 16 [i. In view of 
the fact that although B. satyri differs from the type species, 
J?, studeri, in that the ovarian portion of its vagina is enlarged 
as shown in Chandler's text figure (Chandler, 1925) and that 
there exist marked discrepancies in the disposition of the vagina 
and in the number of testes (only 100 in B. satyri) between the 

8 The Philippine Journal of Science mB 

two, yet the former is being considered a synonym of the latter ; 
we, therefore, prefer to place our material provisionally under the 
type species, B. studeH, until the synonymy of the described spe- 
cies is definitely established. However, we feel justified in fur- 
nishing a detailed description of our material, because after all 
it may prove to be a new species. 

In almost all recorded cases of Bertiella infestation in man, 
the subjects were young children, Mapleston (1930) has made 
the same observation. This phenomenon is shared by other rare 
human tapeworms; such as, Davainea, Hymenolepsis, Dipylidium, 
Drepanidotmnia, and Raiilietina (Faust, 1930), The authors 
(Africa and Garcia, 1934) report two cases of rare human tape- 
worm infestations; namely, (a) a case of a Filipino child two 
years old who passed spontaneously a small tapeworm, which 
they identified as Davsenia madagascariensis, and (6) another 
tapeworm recovered from a hospital patient, a child also two 
years old, which proved to be Raiilietina garrisoni Tubangui, 
1931. The latter worm is a very common intestinal parasite 
of the Philippine brown rat {Mus norvegicus Erxleben, 1777). 
Joyeux and Baer (1929) are of the opinion that Garrison's 
Davaenia madagascariensis and other rare human tapeworms 
are parasites of wild animals and accidentally transmitted to 
man. The fact that in most cases of, rare human tapeworm 
infestations the subjects are young children seems to support 
further the h3rpothesis advanced by Looss (1911:240) and 
shared by Sandground (1929), which seeks to explain the pheno- 
menon of age resistance observed in certain metazoan infesta- 
tions. This hypothesis is that the young of many different spe- 
cies of animals possess certain physiological characteristics in 
common, but as they grow older physiological changes along 
different lines develop, leading to the divergence observed in the 
adult, "In view of this it is reasonable to suppose that the 
closer the genetic relationship between two hosts, the more 
similar will be the conditions in those hosts when young, in so 
far as they will approach the requirements of the parasites for 
their existence. If the premise is not an erroneous one, it is 
not difficult to understand how a parasite, best adapted to live 
in one host, should be able to establish itself for a time in young 
individuals of another somewhat related host, yet fails to ac- 
complish this as the host grows older and acquires specific 
characters that are unfavorable to the parasite." 

The situation developing from the fact that most of the human 
infestations of Bertiella so far recorded have occurred in lands 

66,1 Africa and Garcia: Bertiella 9 

where primates (orang-outang, chimpanzees, and monkeys) 
are in abundance and often in close association with human 
beings, seems to fit the hypothesis expounded above. The si- 
multaneous appearance of B. mucronata in a young Spaniard 
and in the chimpanzees in Cuba (Cram, 1928), our own finding 
(this report) of Bertiella sp. in a child and a monkey in the 
Philippines, and the four reports on children from India and 
the nearby island Sumatra where monkeys thrive in abundance, 
seem to strengthen this hypothesis. 

Exactly four months after receiving our human Bertiella from 
Doctor Gustilo, we received from the same source another spec- 
imen consisting of an apparently complete strobila minus the 
scolex, 20 centimeters long, with two groups of detached ripe 
segments, passed after administration of male fern. It is pre- 
sumed that this specimen developed from the scolex that was left 
behind last April. 


Specimens of Bertiella species from a Filipino child, a dog, 
and a Philippine monkey {Macacus cynomolgus) in the Philippine 
Islands are described. This report provides the tenth and new- 
est case in the list of this rare human-tapeworm infestation, 
establishes a new animal host for the genus, and defines a new 
area in its geographical distribution. A short discussion of the 
relation of tapeworms commonly infesting wild animals to the 
human host is given. 


The writers are indebted to Dr. Marcos Tubangui, of the 
Bureau of Science, Manila, for his courtesy in placing his mon- 
key material at our disposal; and to Dr. Jesus Gustilo, of Sara, 
Iloilo, for cooperating with us by presenting to our laboratory 
specimens of the tapeworm that were passed by his daughter. 


Adams, A. R. D., and Lewis Webb. Two further cases of human infesta- 
tion with Bertiella studeri (Blanchard, 1891) Stiles and Hassal, 1902, 
with some observations on the probable synonymy of the specimens 
previously recorded from man. Ann. Trop. Med. & Parasit. 27 3 

Africa, C. M., and E. Y. Garcia. A rat tapeworm (Raillietina garri- 
soni Tubangui, 1931) transmissible to man, with notes on Davainea 
madagascariensis Garrison, 1911. Philip. Journ. Pub. Health 1 No. 
1 (October, 1934) 44-51. 

10 The Philippine Journal of Science 

Bbddard, Frank E. Contributions to the anatomy and systematic arrange- 
ment of Cestoidea. Proc, Zool. Soc. London, September-December, 
1911 (1911) 622-670, figs. 148-159. 

Blanchard, R. Bertiella satyri, de Torang-outang, est aussi parasite de 
lliomme. Bull. Acad. Med. Paris III 69 (1913) 286. 

Cameron, T. W. M. A new record of the occurrence of a tapeworm of 
the genus Bertiella in man. Journ. Helminth. 7 (1929) 231. Re- 
viewed in Trop. Dis. Bull. 28 (1931) 204. 

Chandler, A. C. New records of Bertiella satyri (Cestoda) in man and 
apes. Parasitol. 17 (1925) 421. 

Cram, Eloise B. A species of the cestode genus Bertiella in man and the 
chimpanzee in Cuba. Am. Journ. Trop. Med. 8 (1928) 339. 

Faust, E. C. Human Helminthology. Lea and Fibeger, Philadelphia 

JoYEUX, Ch., and J. G. Baer. Bulletin de la Societe de Pathologie Exo- 
tique 22 (1929) 114-136. 

Looss, A. The anatomy and life history of Agchylostoma Dub. Rec. Egypt. 
Govt. Med. School, Cairo 4 (1911) 613. 

Maplbstonb, p. a. ;New case of Bertiella studeri in a human being. Ind. 
Med. Gaz. 65 (1930) 258. Reviewed in Trop. Dis. Bull. 28 (1931) 

Meggitt, p. J. Report on a collection of Cestoda, mainly from Egypt. I: 
Families Anophlocephalidae, Davaineidae. Parasit. 19 (1927) 314. 

MUKERJI, A. K. The incidence of helminthic infections in the Carmichad 
Hospital for Tropical Disease, Calcutta. Ind. Med. Gaz. 62 (1927) 
695. Reviewed in Trop. Dis. Bull. 25 (1928) 448. 

Sandground, J. R. A consideration of the relation of the host specificity 
of helminths and other metazoan parasites to the phenomena of age 
resistance and acquired immunity, Parasit. 21 (1929) 227-255. 

Southwell, T. The Fauna of British India, including Ceylon and Burma: 
Cestoda. London 2 (1930). 

Turtox News. Published monthly by the General Biological Supply House, 
761-763 East 69th Place, Chicago. 


Platej 1 

Pig. 1. Photograph of strobila of Bertiella sp. from man. 
. 2. Microphotograph of ovum of Bertiella sp. from man. 
8. Microphotograph of ovum of Bertiella sp. from man, after clear- 
ing with Puri's fluid showing the pyriform apparatus. 

Plate 2 

Fig. 1. Camera-lucida drawing of a mature segment of Bertiella sp. from 
man (lateral view). 

2. Ripe segment of Bertiella sp. from man (lateral view) . 

3. Longitudinal section of a mature segment of the same tapeworm 

just medial to the genital pore. 

4. The same segment at the level of the seminal vesicle. 

6. The same segment as the above but at the level of the ovary. 

6. Ventral view of Bertiella sp. from man, reconstructed from sec- 


7. Camera-lucida drawing of ovum of Bertiella sp. from man. 

Plate 3 

Fig. 1. Scolex of Bertiella sp. from monkey. 

2. Camera-lucida drawing of a mature segment of Bertiella sp. from 

monkey (lateral view). 

3. A transverse section of a mature segment of Bertiella sp. from 


4. Another transverse section of a mature segment of Bertiella sp. 

from monkey. 

5. Ovum of Bertiella sp. from monkey. 

Plate 4 

Fig. 1. Free-hand drawing of a mature segment of Bertiella sp. from dog 
(ventral view) . 
2. Camera-lucida drawing of a transverse section of a mature seg- 
ment of Bertiella sp. from dog. 


< ; %su I N ; r.v-<.-' \)'\ J 

tPHlLlP. JouiN. Sci., 56, No. 1. 



f>-, ^ '^4-, .4i^4< 



■I ^^^ 


AlftK^A ANf iiAltiiA; llhliriKt.LA, | 

[PmUP. JOITIIX. Stl.. H\, Nil 


ArUHH -4NII <;Afi<^IA PH{TfFfJ.\.l 

|l*»iL.|F. ,l(»Mt\. .<ri., r>tH N«». I. 

, . ,. • \,.,j 

t ; . 
■ I .^■ 



1 mm 

•" I, 




i_ -f. 

r/V-y-rl*^// '1-/^/^ 

1 mm 

V" ' '' 













By Marcos A. Tubangui 
Of the Division af Biological ProcLucts, Bureau of Science, Manila 


The four species of Acanthocephala described in this paper 
were collected from bird hosts by Messrs. A. Duyag and P. Ri- 
mando, to whom I wish to express my appreciation. Two of the 
parasites are represented by single specimens, for which reason 
it has proven difficult to make an accurate determination of their 
systematic affinities. For purposes of description they have been 
assigned to the genera Oligoterorhynchus Monticelli, 1914, and 
Prosthorhynckus Kostylev, 1916, as defined by Southwell and 
Macfie (1925). 

Family ECHINORHYNCHID^ Cobbold, 1879 

POLYMORPHUS FRONTOSPINOSUS sp. nov. Plate 1, fi»». 1 to 4. 

Material — Eight males and fourteen females, all mature, from 
the intestine of the common night heron, Nycticorax nycticorax. 

This parasite was in the beginning thought to be identical 
with Arhythmorhynchus hispidus Van Cleave, 1925, a parasite 
of Nycticorax nycticorax in Japan, as described by Fukui (1929) . 
Closer examination, however, revealed that it properly be- 
longs to the genus Polymorphus Luehe, 1911, which, according 
to Travassos (1926), includes three species also parasitic in 
herons of the genus Nycticorax. It presents a great similarity 
to Polymorphus mutdbilis (Rudolphi, 1819) , from which it may 
be distinguished by the larger dimensions of its body and organs, 
the number of its proboscis hooks, and the uniform distribution 
of cuticular spines at the anterior region of its body. 

Description. — Sexual dimorphism not strongly marked. Body 
more or less claviform, being swollen near anterior end and 
gradually tapering towards posterior extremity to a knoblike 
process which is especially noticeable in the female. Male meas- 


14 The Philippine Journal of Science ^^^^ 

ures 9,5 to 12.0 by 1.8 to 2.2 millimeters, female 12.5 to 15.0 
by 2.0 to 2.5 millimeters. Cuticle from anterior end to posterior 
level of proboscis sheath in both sexes covered with numerous 
minute spines 10 to 12 microns long. 

Proboscis subcylindrical, slightly swollen behind middle of its 
length, 0.70 to 0.80 by 0.25 to 0.32 millimeter in male and 0.80 
to 0.92 by 0.30 to 0.34 millimeter in female. Except at its 
base, it is profusely covered with strong hooks arranged radially 
in eighteen to twenty alternating longitudinal rows of fourteen 
to fifteen hooks each. Hooks near middle of proboscis larger 
than those at anterior and posterior ends ; they are 45.7 to 48.9 
microns long and provided with rectangular roots 48.9 to 62.5 
by 16.5 to 20.8 microns. 

Neck absent. 

Proboscis sheath double-walled, that of male measuring 1.28 
to 1.90 by 0.28 to 0.50 millimeters and that of female 2.0 by 
0.36 millimeters. Central nervous system immediately behind 
middle of length of proboscis sheath. Retinacula short, narrow. 

Lemnisci slightly longer than, at most three-halves as long as, 
proboscis sheath. In male they barely reach the anterior level 
of the first testis. 

Testes small, oval, very close together, one obliquely behind 
the other, at junction of first and second thirds of body length; 
they measure 0.42 to 0.90 by 0.36 to 0.70 millimeter. Prostatic 
glands tubular, six in number, compacted into an elongated mass 
2.7 to 4.3 millimeters long. Cement reservoir 1,24 to 1.70 by 
0.50 to 0.54 millimeters. Bursa relatively small. 

Mature eggs in body cavity long, narrow, with three mem- 
branes, the middle one of which is thicker and with an out- 
pocketing at each pole; they measure 104 to 120.6 by 29.1 to 
33.3 microns. 

Host. — Nycticorax nycticorax. 

Location. — Intestine. 

Locality. — Novaliches, Rizal, Luzon. 

Type specimens. — Philippine Bureau of Science parasitological 
collection No. 300. 

MEDIORHYNCHUS SIPOCOTENSIS sp. nor. Plate 2, fiffs. 1 and 2. 

Material. — One adult male and one adult nongravid female. 

This parasite fits well in the genus Mediorhynchus Van Cleave, 
1916, as redefined by Travassos (1924). In so far as the ar- 
rangement of its proboscis hooks in longitudinal rows is con- 
cerned, its nearest ally known is M. oswaldocruzi Travassos, 1923, 

56,1 Tubangui: Acanthocephala 15 

It may be distinguished from the latter by its smaller body 
size, its larger proboscis hooks, and its shorter lemnisci. 

Description. — Sexual dimorphism marked. Body flattened 
laterally, broadest near anterior end ; posterior end in both sexes 
bluntly rounded. Cuticle slightly rugose and presenting indica- 
tions of pseudosegmentation. Male 5.7 millimeters in length 
by 1.2 millimeters in maximum diameter, female 12.5 by 2.2 

Proboscis cylindrical to subcylindrical, truncated anteriorly, 
measures 0.42 by 0.32 millimeter in male and 0.52 by 0.46 
millimeter in female; it is armed with prominent hooks ar- 
ranged in twenty alternating longitudinal series of five or six 
hooks each. The hooks protrude through the centers of papil- 
liform projections, the presence of which give the surface of the 
proboscis a rugged appearance. In the male the hooks (lamina) 
are 50 to 62.5 microns long and have rectangular roots 50 to 
58.2 by 16.5 to 18.7 microns; in the female the lamina of the 
hooks are 62.4 to 66.5 microns in length and the roots 54 to 
62.5 by 16.5 to 20.8 microns. 

Neck well defined and like that of the proboscis its surface 
is rugged due to the presence of papilliform tubercles, through 
the centers of which small hooks protrude ; in the male it meas- 
ures 0.24 by 0.32 millimeter and in the female 0.30 by 0.60 
millimeter. The hooks are arranged irregularly in about forty 
longitudinal rows of five or six hooks each; they are without 
roots and in the male they measure about 20 microns and in the 
female 25 to 33.3 microns in length. 

Proboscis sheath poorly developed, single-walled, measures 
1.04 by 0.35 millimeters in male and 3.03 by 0.82 millimeters in 
female. Central nervous system in middle of length of probos- 
cis sheath. 

Lemnisci of male 3.9 millimeters long by 0.2 millimeter in 
maximum diameter, those of female 6.2 by 0.2 millimeters. 

Testes oval, tandem, partly overlapping, in third fourth of 
body length; anterior testis 0.98 by 0.38, posterior testis 0.76 
by 0.42 millimeter. Prostatic glands eight in number, roundish, 
0.10 to 0.12 millimeter in diameter, immediately behind testes. 
Cement reservoir 0.34 by 0.28 millimeter. Bursa small, in- 

Body cavity of female devoid of ova. 

Host. — Penelopides manillm. 

Location.— Intestine. 

16 The Philippine Journal of Science int 

Locality. — Sipocot, Camarines Sur, Luzon. 
Type specimens. — Philippine Bureau of Science parasitological 
collection No. 332. 

PBOSTHOBHYNCHUS PITTARUM sp. noT. Plate 2, fiars. S and 4. 

Material. — One adult female. 

Description. — Body plump, tapering towards both extremities, 
about 10 millimeters in length by 2 millimeters in maximum 
diameter; posterior end rounded, slightly curved ventrally. 
Cuticle unarmed. 

Proboscis cylindrical, rounded at anterior end, 0.75 by 0.20 
millimeter, armed with fourteen longitudinal rows of fifteen 
hooks each. Hooks 54 to 58 microns long and provided with 
roots 41 to 45 microns long. 

Neck short, 0.12 millimeter long, unarmed. 

Proboscis sheath with double walls, 1.6 by 0.4 millimeters. 
Central nervous system not evident. 

Lemnisci narrow, 1.95 millimeters long. 

Body cavity filled with numerous eggs, the apparently mature 
ones provided with three concentric membranes and measuring 
125 to 130 by 45 to 50 microns. Genital opening ventrosub- 
terminal at posterior end. 

Host. — Pitta atricapilla. 

Location. — Intestine. 

Locality. — Novaliches, Rizal, Luzon. 

Type specimen. — Philippine Bureau of Science parasitological 
collection No. 384, 


Material. — One adult male. 

Description. — ^Body quite slender, cylindrical, about 10.5 by 
0.9 millimeters; posterior end bluntly conical. Cuticle smooth. 

Proboscis subcylindrical, rounded at anterior end, 1.1 by 0.22 
millimeters ; base unarmed, the rest of its surface profusely 
covered with prominent hooks arranged radially in eighteen 
longitudinal series of twenty-one hooks each. Hooks on ex- 
treme anterior and posterior ends of proboscis smaller, with 
reduced roots, and 33.3 to 41.5 microns long; rest of hooks 41.5 
to 49.5 microns long and with well-developed roots 37.5 to 41.6 
by 16.5 microns. 

Neck absent. 

Proboscis sheath double-walled, 1.84 by 0.35 millimeters. 
Central nervous system in front of middle of length of pro- 
boscis sheath. Retinacula narrow, short. 

96,1 Tvhangui: Acanthocephala 17 

Lemnisci narrow, 2.3 millimeters long. 

Testes oval, one immediately behind the other, in middle of 
body length; first testis 0.85 by 0.40, second testis 0.88 by 0.36 
millimeter. Prostatic glands four in number, tubular, forming 
a bundle 2.7 millimeters long. Cement reservoir 1.3 by 0.35 
millimeters. Bursa well developed, about 1 millimeter across, 
arising from ventral aspect of posterior end of body. 

Host. — Hypotsenidia pkilippensis. 

Location. — Intestine. 

Locality. — Novaliches, Rizal, Luzon. 

Type specimen. — Philippine Bureau of Science parasitolog- 
ical collection No. 380. 


FUKUI, Tamao. On some Acanthocephala found in Japan. Annot. Zool. 

Japon. 12 (1929) 255-270. 
Southwell, T., and J. W. S. Macfie. On a collection of Acanthocephala 

in the Liverpool School of Tropical Medicine. Ann. Trop. Med. and 

Parasit. 19 (1925) 141-184. 
Travassos, L. Contribucoes para o conhecimento da fauna helminthologica 

brazileira. XX. Revisao dos Aconthocephalos brazileiros. Parte II. 

Familia Echinorhynchidae Hamann, 1892, sub-fam. Centrorhynchinae 

Travassos, 1919. Mem. Inst. Oswaldo Cruz 19 (1926) 31-125. 
Van Cleave, H. J. Acanthocephala of the genera Centrorhynchus and Me- 

diorhynchus (new genus) from North American birds. Trans. Am. 

Micros. Soc. 35 (1916) 221-232. 



[Drawn t>y Alfredo C. Gonzales.] 

Plate 1 
polymorphus frontospinosus sp. nov. 

Fig. 1. Male, ventrolateral view. 

2. Female, ventrolateral view. 

3. Proboscis hooks. 

4. Mature egg. 

Plate 2 

Fig. 1. Mediorhynchus sipocotensis sp. nov.; male, lateral view. 

2. Mediorhynchus sipocotensis sp. nov.; proboscis of female, lateral 


3. Prosthorhynchus pittarum sp. nov.; female, lateral view. 

4. Egg of Prosthorhynchus pittarum sp. nov. 

5. Oligoterorhynchus malayensis sp. nov.; male, lateral view. 



CPhilip. Journ. Sci., 56, No. 1. 


0.05 mrtv 







[Philip. Journ. Scl, 56, No. 1. 



Of the Division of Biological Products, Bureau of Science, Manila 


It is a welWcnown fact that beriberi occurs in rice eaters and 
is the result of some deficiency in the diet; namely, lack of 
vitamin B. 

Judging from the report of the Bureau of Health, beriberi 
is a great public-health problem in the Philippine Islands, for 
although it is decreasing in Manila, it is increasing in the pro- 
vinces, so that the total number of cases increases every year. 

By courtesy of Prof. Luis Guerrero I was able to examine 
a few cases of beriberi confined in the Philippine General Hos- 
pital. It was found that the diastolic blood pressure in these 
cases was high compared with that observed in cases of beriberi 
in Japan, and the question arose whether or not race, climate, 
and food might explain this apparent difference. An examina- 
tion was therefore undertaken of the clinical case records of 
beriberi cases admitted to the Philippine General Hospital during 
the period from May, 1932, to June, 1934, in order to determine 
if there actually existed any difference between the diastolic 
blood pressure of beriberic Filipinos and beriberic Japanese. 


Beriberi is a composite of many symptoms, such as palpitation, 
oedema, digestive, sensory, and motor disturbances, etc., and re- 
cently Z. Shimazono(i) and K. Omori(2) have reported on the 
lowering of the diastolic blood pressure in beriberi compared 
with the normal average. This lowering, according to these 
authors, is also observed in experimental beriberi in human 
beings. Shimazono(i) states that this characteristic symptom 
is observed even in incipient cases of beriberi. 

It has been found that the cause of this lowering of the dias- 
tolic blood pressure is a weakening of the arterial tension, and 



The Philippine Journal of Science 


does not depend upon a diminution in the blood quantity. It 
has also been observed that the tension of the heart and the 
peripheral vascular system is always more or less depressed 
in cases of beriberi. 

Table 1, from a case reported by Shimazono, shows the low- 
ering of the diastolic blood pressure in beriberi. 

Table 1. — Blood pressure of M, N., student, age 21 years {by Riva-Rocci) . 





September 4_, 
September 5_ 
September 7. 
September 10 
September 13 
September 21 
September 25. 
September 80, 
October 5 

mm Hg. 

mm Hg. 





Very serious. 


Very well. 

In the systolic blood pressure beriberi causes no remarkable 
change. However, in serious cases a lowering of the systolic 
blood pressure may be observed, which becomes a fatal omen 
when it drops to 80 or 70 millimeters of mercury, according to 
Riva-Rocci. In some cases recovery from beriberi is associated 
with an increase in blood pressure above normal, reaching 150 
or 160 millimeters of mercury even in young individuals, but 
returning gradually to normal upon recovery of the patient, 
as in the case by Shimazono (Table 1). 

On the blood pressure in beriberi, especially the diastolic blood 
pressure, nothing has been written in the Philippines where 
beriberi is very prevalent. Therefore, it would be interesting 
and pertinent to investigate this problem. 


Much work has been done in the study of blood pressure 
among Filipinos. However, no author except Concepci6n(3) has 
reported anything on the diastolic blood pressure. His work 
has, therefore, proved invaluable to me, supplying me with the 
very much needed standard basis of comparison of diastolic 
blood pressure in Filipinos. 


Sugino: Blood Pressure in Beriberi 


Following Concepcion's classification, (3) i have divided my 
cases into eight groups by age and compared the diastolic blood 
pressure of beriberi with that of average normal readings. 

Ninety-eight cases of beriberi were admitted to the Philippine 
General Hospital during the period from May, 1932, to June, 
1934, of which forty-one, or 42 per cent were males, and fifty- 
seven, or 48 per cent, females. 

Table 2 shows the incidence of beriberi according to decades 
of age, except for the cases between 15 and 20 years of age, 
studied by Concepcion. 

The incidence of beriberi is highest in individuals 21 tb 
30 years of age, especially females. Next in order come cases 
occurring in young people between 15 and 20 years of age, 
after which the tendency decreases with age. Beriberi is a 
disease that, like tuberculosis, is easily contracted by young 
adults between the ages of 21 to 30 years. 

Table 2. — Incidence of beriberi by age and sex. 

Age in years. 


Age In years. 



Per cent. 


Per cent. 

15-20 _ 









21-30 _ _ 












Tables 3 and 4 show the relation between the diastolic blood 
pressure of the beriberi cases that I have collected and the 
corresponding normal average readings as reported by Con- 

Only two male and six female beriberi cases show higher 
diastolic blood pressures than the normal diastolic blood pressures 
reported by Concepci6n. In all other beriberi cases the diastolic 
blood pressure was always below the normal average. These 
findings agree with those of Shimazono and show that race, 
climate, and food, which on first sight seemed to exert some 
influence, actually have no effect. The average diastolic blood 
pressure observed in beriberi cases in the Philippines is the 


The Philippine Journal of Science 


same as that observed in beriberi cases occurring in Japanese 
in Japan. In the eight cases above showing diastolic blood 
pressure higher than normal, the systolic blood pressure was 
likewise remarkably higher than normal (Tables 3 and 4). 

Table 3. — Blood pressure 

in normal and in beriberic male Filipinos. 

Age in years. 









mm Hg. 

mm Hg, 

' 118.0 

' 70.0 







16-20. .- 



- 77.6 








. 50.0 

' 90.0 

" 60.0 





b 132.0 






- 108.0 

. 76.6 








. 118.0 

. 80.0 


" 70.0 







^ 115.0 

. 80.8 


b 140.0 


. 85.0 

. 42.0 

' 115.0 







41-60 - 





. <•) 





. 118.0 

. 80.0 

g j^.gO 


f 135.0 
\ 130.0 

I 86.2 

[ 80.0 
i 60.0 

- 125.0 

. 70.0 

gj^-go _ - . 




J 90.0 

95 .0 
. 170.0 

. 90.0 

» No record. 

b Much higher than that of Concepci6n. 

56,1 Sugino: Blood Pressure in Beriberi 25 

Table 4. — Blood pressure in normal and in beriberic female Filipinos. 

Age in years. 








mm Hg. 

mm Hg. 

' 110.0 

' 46.0 

b 140.0 




15-20 -- 



. 80.8 






\ 98.0 

. 60.0 

' 110.0 

' 68.0 









b 142.0 









b 140.0 











" 104.0 


, 84.1 


- 86.0 


























. 105.0 


- 118.0 

' 70.0 





3l~40 , 



" 85.1 - 






. 113.0 . 


a No record. 

bMuch higher than that of Concepcidn. 

26 The Philippine Journal of Science 1935 

Table 4. — Blood pressure in normal and in beriberic female Filipinos — Ctd. 

Age in years. 









41-60 __- - --- -- - 


mm Hg. 

' 120.0 

' 90.5 

mm Hg. 

Tables 3 and 4 show that the difference between the systolic 
blood pressure of the above-excepted eight beriberi cases and 
the corresponding normal systolic average is greater than the 
difference existing between the diastolic blood pressure of these 
eight cases and the corresponding normal diastolic average given 
by Concepci6n, which further shows that there actually occurs 
an absolute lowering of the diastolic blood pressure in beriberi. 


1. The incidence of beriberi is highest in individuals ranging 
from 21 to 30 years of age, especially females, followed closely 
by young adults of 15 to 20 years. With advancing age there 
is a tendency for beriberi to decrease. 

2. The diastolic blood pressure in beriberi is lower than the 
normal average diastolic blood pressure at a given age. 

3. When the systolic blood pressure is higher than normal in 
beriberi, the lowering of the diastolic blood pressure is not evi- 


In conclusion I wish to express my obligation to Prof. L. 
Guerrero, of the Philippine General Hospital, and Dr. M. Tu- 
bangui, of the Bureau of Science, for permission to carry on 
these studies. 

56,1 Sugino: Blood Pressure in Beriberi 27 


1. Shimazono, Zunjiro. Kakke (in Japanese) (1929) 97-100. 

2. Omori, Kenta. Kakke (in Japanese) (1927). 

3. CoNCEPCiON, ISABELO, and Emilio Bulatao. Philip. Journ. Sci. § B 3 

(1916) 135-149. 

4. Lantin, Gregorio T. Journ. Philip. Is. Med. Assoc. (1938) 191-195. 

5. Chamberlain, Weston P. Philip. Journ. Sci. § B 6 (1911) 469. 

6. Musgrave, W. E., and A. G. Sison. Philip. Journ. Sci. § B 5 (1910) 


7. Vedder, Edward B. Beriberi (1913) 21-54. 


By GuiLiiERMo J. Blanco 

Of the Fish and Game Administration, Bureau of Science, ManUa 


Fresh-water units of the Philippine Islands are generally 
richly inhabited by small prawns locally called apta, yapyap 
(Tagalog), daliw-daliw, or koros (Ilocano). These prawns be- 
long to the family Atyidse. They are abundant in large fresh- 
water lakes, especially Laguna de Bay and Taal Lake, wher^ 
they are caught in large quantities by means of scissors-nets, 
such as the salap and sakdg. 

This crustacean is eaten fresh, or salted and made into a fer- 
mented product called alamang. It is also simply dried and 
sold as dry prawn. When the supply is abundant, it is prepared 
as protein feed for ducks and chickens or converted into some 
form of fertilizer. The local price of this product varies from 
1 peso (50 cents United States currency) to 2.50 pesos a ca- 

Although these prawns are widely distributed in the Tropics 
and Subtropics, very little has been written on their distribution 
and systematics in the Philippines. R. P. Cowles, formerly of 
the Department of Zoology, College of Liberal Arts, Univer- 
sity of the Philippines, described the feeding habits of Atya 
moluccensis de Haan, so far the only described species in the 
Philippines. He indicated that there are several undescribed 
species of Caridina, which have feeding habits more or less sim- 
ilar to those of Atya moluccensis. 


In general, the iRrst and second pairs of legs in the Atyidse 
are comparatively small and of similar size. The chete are 
each fringed with a brush of long hairs. The family is rep- 
resented in the Philippine Islands by Atya, Caridina, and Ort- 
mania, which are believed to be more specialized than genera 
occurring in other countries. 

^ A cavan is 25 gantas. A ganta is equivalent to 3 liters or 3.3036 quarts. 


30 The Philippine Journal of Science i^^s 

The family Atyidse is represented in the collection of the 
Bureau of Science by the following: 

Atya moluccensis de Haan. 

Atya serrata Spence Bate. 

^QtHdina gradllima Lanchester. 

Caridina gracilirostris de Man. 

Ccmdiim nilotica var, brachydactyla de Man. 

Caridina modigliani Nobili. 

Caridina brevicarpalis var. endehensis de Man. 

Caridina laevis Heller. 

Ortmania sp. (?). 

Genus ATYA Leach 

Atya Leach, Trans. Linn. Soc. London 11 (1815) 345. 

Pairs of chelipeds quite similar; carpus reduced by the exca- 
vation of its distal border to a narrow crescent form ; propodus 
modified, dactylus markedly developed so that the "palm" is 
entirely absent; chela composed of two similar parts. 

Key to the known species of Atya in the Philippines,, 

a\ Infraorbital and pterygostimal angles of anterior carapace prolonged 
to a point A, moluccensis de Haan. 

a^. Infraorbital and pterygostimal angles of anterior carapace not long 
and pointed A, serrata Spence Bate. 


Atya moluccensis de Haan, Fauna Japonica, Crustacea (1849) 184- 

186, pi. 21; Miers, Ann. & Mag. Nat. Hist. (V) 5 (1880) 42, pi. 

15, figs. 3 and 4; de Man, Zool. Ergeb. (1892) 357, pi. 21, fig. 20; 

Ortman, Denk. Ges. Jena 8 (1894) 1-80, pis. 1-13; Proc. Acad. 

Nat. Sci. Phila. (1895) 408; de Man, Abh. Senck. Naturf. Ges. 138 

(1904) 137. 
Atya armata Milne-Edwards, Ann. Soc. Ent. de France (1864) 149, 

pi. 3, fig. 3; von Martens, Arch. Naturg. Jahrg. 47 (1868) 47, pi. 

1, ^g. 6. 
Atya gustavi Ortman, Zool. Jahrb. Syst. 5 (1890) 467, pi. 36, figs. 

a, b, c. 
Atya dentirostris Thallwitz, Abh. Ber, K. Zool. Anthr. ethn. Mus, 

Dresden (1891) 26, fig. 7. 

Rostrum short, narrow, slenderly excavated at broad side of 
dorsal part and armed with numerous little teeth on ventral 
keel. Infraorbital and pterygostimal angles prolonged to a 
point, "Immobile" and "mobile" fingers of first and second 
peraeopods alike in shape and size. Excavation of distal borders 
of carpus crescent-shaped. 

Three adult specimens, 35 to 39 millimeters long. 

^^'^ Blanco: Philippine Atyidse 31 

l^EGROS, Oriental Negros Province, Dumaguete River, March 

ATYA S^RRATA Spence Bate. Plate 1, figs. 1 to 4. 

, Atya serrata Spence Bate, Challenger Kept. Zool. 24 (1888) 669, pi. 
119, %. 2; Ortman, Proc. Acad. Nat. Sci. Phila. (1895) 410; 
Bouvier, C. E. Acad. Sci. 138 (1904) 446; Bordage, C. R Acad 
Sci. 147 (1908) 1418, fig. 1; 148 (1909) 47; Caiman, Quart. Jcmm! 
Micr. Sci. 55 (1910) 792; Bouvier, C. R. Acad. Sci. 152 (1911) 
1822; 154 (1912) 692, figs. 4, 6, 7; Trans. Linn. Soc. London (II) 
15 (1913) 460; C. R. Acad. Sci. 159 (1914) 700; Bouvier end 
d'Emmeres de Charmoy, C. R. Acad. Sci. 169 (1919) 317. 

Ortmania alluaudi mutation serrata Bouvier, Bull. Scient. de Fr. et 
Belg. 39 (1905) 106, 115, fig. 19. 

Atya hrevirostris de Man, Zool. Ergeb. 2 (1892) 360, pi. 21, fig. 21; 
Ortman, Denk. Ges. Jena 8 (1894) 10; Ortman, Proc. Acad. Nat, 
Sci. Phila. (1895) 409; Schenkel, Verh. Naturf. Ges. Basel 13 
(1902) 500, fig. 6; Bouvier, Bull, du Mus. (1904) 137. 

Atya breverostris var. de mani Nobili, Ann. Mus. Civ. dei Storia 
Nat. Genova (II) 20 (1900) 475. 

Atyoida tahitensis Stimpson, Proc. Acad. Nat. Sci. Phila. (1860) 97. 

Infraorbital and pterygostimal angles of anterior part of 
carapace very short and pointed. Short rostrum not extending 
beyond second segment of antennular peduncle, curved down- 
wards, ventrally forming a keel. Formula of rostral teeth |. 
'Immobile" and "mobile'' fingers alike in shape and size in 
both first and second perseopods. Excavations of carpus of 
their distal borders, crescent-shaped. Apex of telson semicir- 
cular, possessing two externolateral spines and two interno- 
lateral spines. 

Several specimens, 18 to 29 millimeters long. 

Luzon, Ilocos Norte Province, Laoag River, February 5, 1934. 

Genus CARIDINA Milne-Edwards 

Caridina Milne-Edwards, Histoire naturelle des Crustaces 2 (1837). 

Chelae not very variable ; dactylus or "movable finger" opposed 
to "immobile finger" or propodus ; carpus of first pair short and 
broad, its distal margin more or less concave ; propodus articulat- 
ing with lower corner of carpus; carpus of second pair more or 
less elongated and slender; propodus articulating with its distal 

Key to the knovm species of Caridina in the Philippines. 

a\ Rostrum long, curving upward towards tip; end bifid or trifid. 


6*. Number of teeth of upper and lower borders of rostrum — . 


C gracillima Lanchester. 

32 The Philippine Journal of Science im8 

b*. iKumber of teeth of upper and lower borders of rostrum yarjriiiff 

between - ^ -^ - C. gracUiroBtf%» de Man. 

h *. Number of teeth of upper and lower borders of rostrum yarying 

from ^t:^ C. modigliani Nobfli. 


b \ Number of teeth of upper and lower borders of rostrum varying from 


. C. brevicarpalis var. endehensis de Man. 

o*. Rostrum short, not extending beyond antennal scale. 

6 *. Number of teeth of upper and lower border of rostrum -~ ; end 

trifid C. nilotica var. endehensis de Man. 

b \ Number of teeth of upper and lower borders of rostrum varying from 

ltl2 C. Isevis Heller. 


CARDINA GRACILLIMA Lanchester. Plate 1, figs. 5 to 10. 

Caridina gracillima Lanchester, Proc. Zool. Soc. London (1901) 560- 
563, pi. 34, fig. 1; Bouvier, Bull. Sclent, de Fr. et Belg. 39 
(1905) 72; C. R. Acad. Sci. 154 (1912) 918; Trans. Linn. Soc. 
London (II) 15 (1913) 463; Kemp, Mem. Asiat. Soc. Bengal 6 
(1918) 285. 

Rostral formula --p; 19 teeth present on upper edge; lower 

edge bearing 15 teeth. Rostrum long, curving upwardly beyond 
antennal scale. Toothed part of upper edge, one-half less than 
total length. Carpus of first leg, twice as long as broad ; that of 
second 4f times as long as broad. Dactylus of third pair with 
10 spines; that of fifth pair, 60 spines. Uropods each with 12 

A single female specimen, 34 millimeters long. 

Luzon, Ilocos Norte Province, Laoag, Caaoacan River, August 
14, 1933. 

CABIDINA GRACILIROSTRIS de Man. Plate 2, fiss. 11 to 17. 

Caridina gradlirostria de Man, Zool. Ergeb. (1892) 399, pi. 25, fi% 
31; Zool. Jahrb, Syst. 9 (1897) 726; Nobili, Ann. Mns. Civ. St. 
iNat. Genova II 20 (1900) 477; J. Roux, Revue Suisse de Zool. 
12 (1804) 555; Bouvier, BuH. Scient. de Fr. et Belg. 39 (1905) 72; 
Trans. Linn. Soc. London II 15 (1913) 463; Kemp, Mem. Asiat. 
Soc. Bengal 6 (1818) 282. 

Rostrum very long, extending beyond antennal scale. Tooth 
formula \.„jLir ; upper edge armed with 6 to 9 teeth, two of 

them on carapace, in addition to a subapical tooth; lower edge 
with from 17 to 27 teeth. Toothed part of upper edge one- 
fourth less than length of unarmed part. 


Blanea: Philippine AtyiAm 


Tablb h-^Length and rostral formula of young Caridina graeUirostria de 




Specimen — 










2 _ 












4 _ 






6 _- __.- 






6 --- 








8 + 1 

17 _ 



8 _ 



18 _ 








8 + 1 




Carpus of first leg If times as long as broad. Dactylus of 
third pair with 12 spines ; that of fifth pair, 50 spines. Terminal 
joints of third maxilliped with 17 spines. Uropods each with 
9 spines. 

Nineteen young specimens, 20 to 37 millimeters long (Table 1) . 

Luzon, Laguna de Bay, November, 1929. 

CABIDINA NILOTICA var. BRACHYDACTYLA de Man. Plate 2, 1kg, 18. 

Caridina wyckii de Man, Zool. Ergeb. (1892) 386-393, pi. 24, fig. 29, 

cc, dd, f , g, i, ii, k. 
Caridina wyckii var. paucipara Bouvier, Bull. Sclent, de Fr. et Belg. 

39 (1905) 79. 
Caridina nilotica var. hrackydactyla de Man, Rec. Ind. Mus. 2 (1908) 

269, pi. 20, fig. 8. 
Caridina hrackydactyla Bouvier, Trans. Linn. Soc. London II 15 

(1913) 463. 
Caridina hrackydactyla subsp. peninsularis Kemp, Mem. Asiat. Soc. 

Bengal 6 (1918) 270, fig. 10. 

Rostral formula ^ ; 31 teeth present on almost entire length 

of upper edge; 5 teeth on carapace; that of lower edge with 13 

289095 » 

34 The Philippine Journal of Science wss 

teeth; its eiid trifid. Rostrum short, extending a little beyond 
antennal scale. 

Fingers of first pair 2 to 2.5 times as long as palm. Carpus 
of second pair of legs very slender; 6 times as long as broad 

Eggs very numerous and small, 0.39 to 0.44 millimeter long 
and 0.22 to 0.25 millimeter wide. 

A single sexually mature specimen, 27 millimeters long. 

Luzon, Ilocos Norte Province, Laoag, Caaoacan River, August 
14, 1933. 

CARIDINA M0DI6LIANI NobUi. Plate 2, flffs. 19 to 24. 

Caridina modigliani Nobili, Ann. Mus. Civ. Stor. Nat. Genova 20 
(1900) 477; Bouvier, BuH. Scient. de Fr. et Belg. 39 (1905) 72. 

Rostrum slender, curving upwardly towards the tip; very 

long, without spines on almost half of its distal length. Rostral 

- , (13-24) ±2-3 io 4- o>i 4. +1. ^u 

formula tt-te J 13 to 24 teeth on the upper margin 

11— io 

of the rostrum, 2 of them on the carapace; that of the lower 

margin with 11 to 16 teeth; its end bifid or trifid. 

Carpus of the first perseopod. If times as long as broad; that 
of the second 5i times as long as broad. 

Forty-one adult specimens, 23 to 34 millimeters long (Table 2) . 

Luzon, Ilocos Norte Province, Laoag, Laoag River. 


Caridina hrevicarpalis var. endehenaia de Man, Zool. Niederlandisch 
Ost. Indien (1892) 399, pi. 24, figs. 30c and 30e. 

Rostral formula of two specimens :^ r^, approaching 

lo ly 

rostral formula of /^^ given by de Man. Rostrum long, ex- 
tending beyond antennal scale. 

Four mature females, 25 to 29 miUimeters long. 

Luzon, Batangas Province, Pancipit River, April 19, 1934. 

CARIDINA L^VIS HeUer. Plate 3, fiffs. 2$ to 32. 

Caridina Imvis Heller, Verh. Zool. bot. Ges. Wien. 12 (1862) 411; 
de Man, Zool. Ergeb. (1892) 376, pL 22, fig. 27; de Man, Rec. Ind. 
Mus. 2 (1908) 253, pi. 20; Ortman, Proc. Acad. Nat. Sci. Phila. 
(1895) 404; Bouvier, BuU. du Mus. (1904) 131; Bouvier, Trans. 
Linn. Soc. London II 15 (1913) 464; Kemp, Mem. Asiat. Soc. 
Bengal 6 (1819) 289. 

56, 1 

Blanco: Philippine Atyidm 


Rostral formula -^-^ ; upper margin of rostrum with 15 

to 19 teeth, 2 to 5 located on carapace ; lower margin with from 
2 to 6 teeth; rostrum short, not extending beyond antennal scale. 

Tablb 2. — Length and rostral formula of female Caridina modigliani 
























































Specimen — 
























21 + 2 





















The Philippine Journal of Science 


Fifteen young* specimens, 15 to 19 millimeters long (Table 3). 
Luzon, Albay Province, Pulangue Lake, January 26, 1926. 

Table 3. — Length and rostral formula of Caridina lavis Heller, 




Specimen — 



1 _ - . 




































13 -- 

6 - 





Genus ORTMANIA Rathbun 

Ortmania Rathbun, U. S. Fish. Comm. 2 (1910) 120. 
Carpus of first pair of legs similar to that of second pair, being 
short and broad, distal margin excavated and articulating with 
propodus of lower corner. Some species of Ortmania have sec- 
ond carpus similar to that in Caridina, by being somewhat longer 
than the first; excavation of distal margin shallow. Some with 
carpus nearly similar in the two pairs, and deeply excavated to 
assume a crescentic form as in Atya. In some species chelse 
quite similar to those of Caridina, in others "palm" much short- 
ened, with articulation of "movable finger'' carried backward^ 
towards base of propodus. 

ORTMANIA sp. <?). Plate 3, figs. 3a to 40. 

Rostrum short, not extending beyond antennal scale. Rostral 

toothing 5; no teeth on upper edge, and 3 teeth at lower edge. 

Shape of rostrum similar to that of O. meMcana de Saussure. 
Segments of antennular peduncle with numerous spines. 

First perseopod similar in shape to that of Atya. Dactylus of 
third pair of legs with 4 spines; that of fifth pair with 20 spines. 
Apex of telson similar to that of O. alltumdi Bouvier. One of 
the uropods with 15 spines. 

*^' ^ Blanco: Philippine Atyidse 37 

A single young specimen, 14 millimeters long. 

Luzon, Ilocos Norte Province, Laoag, Caaoacan River, August 
14, 1933. 

This specimen is probably a variety of 0. mexicana de Saus- 
sure, if the character of the rostrum is taken as a basis for classi- 
fication, but due to lack of material I hesitate to assign it to 0. 


Annandalb, N., and St. Kemp. The Crustacea decapoda of the Lake Tibe- 
rias. Proc. Asiat. Soc. Bengal (New Series) 9 (1913) 242-253, pis. 

BoRRABAiLE, L. A. On some crustaceans from the South Pacific. Proc. 

Zool. Soc. London (3) (1898) 1000-1005, pis. 63-65. 
BouviER, E. L. Recherches sur la morphologic des variations. La dis- 
tribution geographique des crevettes de la famille des atyides. Ency- 

clopedie Entomologique (1925). 
Calman, W. T. On two species of macrurous crustaceans from Lake 

Tanganyika. Proc. Zool. Soc. London (1899) 704-712, pis. 39, 40. 
Calman, W. T. Zoological results of the Third Tanganyika Expedition. 

Report on the macrurous Crustacea. Proc. Zool. Soc. London. (1906) 

187-206 pis. 11-14. 
Calman, W. T. The researches of Bouvier and Bordage on mutations 

in Crustacea of the family Atyidae. Quart. Journ. Mic. Sc. 55 (1910) 

CowLES, R. P. The habits of some tropical Crustacea, II. Philip. Journ. 

Sci. § D 10 (1915) 11-16, pis. 1-3. 
Dana, J. P. Crustacea U. S. Explor. Exped. (1) 13 (1852) 531-541, pi. 

DE Man, J. G. Decapoden des indischen Archipels. Zool. Ergeb. (2) 

(1892) 295-527, pis. 15-29. 
Haan, W. de. Fauna Japonica de von Siebold. Crustacea (1850) 184- 

186, pi. 21. 
Heller, C. Neue Crustacea gesammelt wahrend der Welt-Umseglung der 

K. K. Pregatte Novara, Verh. Zool.-bot. Gesel. Wien (2) (1862) 525. 
HiCKSON, S. J. On a new species of the genus Atya (A. wyckii) from 

Celebes. Ann. Mag. Nat. Hist. (2) 6 (1888) 357-367, pis. 12-14. 
Kemp, St. Fauna of the Inle Lake. Crustacea Decapoda of the Inle 

Lake Basin. Rec. Ind. Museum 14 (1918) 81-102, pis. 24, 25. 
Lanchester, W. p. On some malacostracous crustaceans from Malaysia 

in the collection of the Sarawak Museum. Ann. and Mag. Hist. (7) 

6 (1900) 249-265, pi. 12. 
Lanchester, W. E. On the Crustacea collected during the Skeat Ex- 
pedition to the Malay Peninsula. Proc. Zool. Soc. London (1901) 533- 

573, pis. 33, 34. 
Spence Bate, C. On some new Australian species of Crustacea. Proc. 
Zool. Soc. London (1863) 498-505. Crustacea Macrura. Challenger. 
Zool. 24 (1889) 691-723. 


Plate 1 
Fig. 1. Atya serrata, anterior cephalothorax. 

2. Atya serrata, perasopod of first pair. 

3. Atya serrata, perseopod of second pair. 

4. Atya serrata, apex of telson. 

5. Caridina gracillimaf rostrum. 

6. Caridina gracillimay perseopod of first pair. 

7. Caridina gracillima, perasopod of second pair. 

8. Caridina gracillima, dactylus of third pair. 

9. Caridina gracillima, dactylus of fifth pair. 
10. Caridina gracillima, uropodial spines. 

Plate 2 
Fig. 11. Caridina gracilirostris, rostrum. 

12. Caridina gracilirostris peraaopod of first pair. 

13. Caridina gracilirostris, perseopod of second pair. 

14. Caridina gracilirostris, dactylus of third pair. 

15. Caridina gracilirostris, dactylus of fifth pair. 

16. Caridina gracilirostris, terminal joint of third maxilliped. 

17. Caridina gracilirostris, uropodial spines. 

18. Caridina nilotica var. brachydactyla, rostrum. 

19. Caridina modigliani, rostrum. 

20. Caridina modigliani, peraBopod of first pair. 

21. Caridina modigliani, peraeopod of second pair. 

22. Caridina modigliani, dactylus of third pair. 

23. Caridina modigliani, uropodial spines, 

24. Caridina modigliani, apex of telson. 

25. Caridina hrevicarpalis var. endehensis, rostrum. 

Plate 3 
Fig. 26. Caridina laevis, rostrum. 

27. Caridina laevis, peraeopod of first pair. 

28. Caridina laevis, peraeopod of second pair. 

29. Caridina laevis, dactylus of third pair. 

30. Caridina laevis, terminal joint of third maxilliped. 

31. Caridina laevis, uropodial spines. 

32. Caridina laevis, apex of telson. 

33. Ortmania sp?, rostrum. 

84. Ortmania sp?, antenna! scale. 

35. Ortmania sp?, antennule. 

36. Ortmania sp?, peraeopod of first pair. 

37. Ortmania sp?, dactylus of third pair. 

38. Ortmania sp?, dactylus of fifth pair. 

39. Ortmania sp?, uropodial spines. 

40. Ortmania sp?, apex of telson. 


Blanco: Philippine Atyid^.] 

[Philip. Journ. Scl, 56, No, 1. 



Blanco: Philippine Atyid^.] 

[Philip. Journ. Scl, 56, No. 1. 


Blanco: Philippine Atyid^:.] 


[Philip. Journ. Scl, 56. No. 1. 






By GuiLLERMo J. Blanco 

Of the Fish and Game Administration, Bureau of Science, Manila 


The caudal fin of fishes is built upon the modification of the 
protocercal embryonic fin. Agassiz, Heckel, Huxley, KoUiker, 
Ryder, and many others have shown that the structure of the 
homocercal fin of bony fishes is a modified structure of the hete- 
rocercal caudal fin. 

The structure of the acanthopterygian caudal fin in six genera 
was studied by Whitehouse (1910) . He found that the structure 
of the tail fin may vary considerably in the same suborder. Ap- 
parently no one has studied the development of the caudal fin in 
the viviparous perches, hence the descriptions of the caudal fin 
of the blue perch are here given. 

The protocercal caudal of a larva 3 mm long (Plate 1, fig. 1), 
is an extension of the notochord, the narrow dorsal fin is not 
continuous with the ventral fin fold. 

The larva when 7 mm long (Plate 1, fig. 2) has an axial lobe, 
axl, that is narrow dorsally and wide ventrally with reference 
to the notochord, which curves slightly downwards. 

The larva when 12 mm long (Plate 1, fig. 3) has a caudal fin 
heterocercal in general form. The notochord has its characteris- 
tic vacuolated structure and is turned upwards in the axial lobe. 
There are traces of six cartilaginous areas, a, b, c, d, e, and /, 
the anlage of the hypurals. Cartilaginous areas d and e have a 
common base, while areas a, b, c, and / are separated. To each 
cartilaginous hypural one or more caudal rays are attached for 
the support of the enlarging membrane of the caudal fin. 

The caudal fin of a larva 14 mm long (Plate 1 fig. 4) has the 
notochord turned upwards toward the axial lobe. The begin- 
nings of the neural spines, ep, first appear above the notochord. 

*This work was done at the University of Washington as a partial 
fulfillment of the requirements for the degree of Master of Science. 


42 The Philippine Journal of Science isss 

The cartilaginous hypurals are still asymmetrically placed, as in 
Plate 1, fig. 3. The caudal fin has grown larger and most of the 
caudal rays arising at the distal end of the hypurals now lie more 
or less in line with the chordal axis. 

The internal parts of the caudal fin of a larva 16 mm long 
(Plate 1, fig. 5) are still asymmetrically placed. Areas a and 
d have assumed a triangular shape; areas 6 and c are now 
parallel; c has its bases fused with those of d and e, so that 
all these hypurals now have a common base, while areas a, &, 
and / are separate. An additional cartilaginous hypural, g, is 
in evidence at this stage. The last two neural spines are also 

Plate 1, fig. 6, shows a caudal fin of a larger larva. The axial 
lobe which was evident in smaller fishes has disappeared in the 
former stages. 

The caudal fin of a larva 21 mm long (Plate 1, fig. 7) shows 
a marked shortening and thickening of the notochord, the hind- 
most portion of which becomes the urostyle. The vertebral 
segments of the caudal fin are still indistinct. Hypurals a, b, f, 
and g are separated, b, c, d, and e have fused bases. Cartilag- 
inous areas a to g are in the process of ossification with the 
ventral part of the chordal axis. The dorsal caudal radials, dcr, 
are between the urostyle and the last neural spine. The caudal 
fin at this stage is externally symmetrical and fan-shaped, but 
internally it is asymmetrical. 

In a larva 41 mm long (fig. 8) hypurals a, b, c, and d are 
joined to the urostyle. Hypural e is connected with the last 
vertebra and g with the penultimate vertebra. The three dorsal 
caudal radials are inserted deeper, and between the last neural 
spine and the urostyle. The segmented dermotrichia, der, are 
now well developed externally and overlap the edge of the three 
dorsal caudal radials, dcr, the hypurals, hy. 

Plate 1, fig. 9, shows the caudal fin of a young perch, 51 mm 
long, in which the third dorsal caudal ray is inserted between 
the urostyle and the first hypural plate. 

Plate 1, fig. 10, is the homocercal caudal of an adult female 
specimen 21 cm long. It has a much reduced urostyle. The last 
dorsal caudal ray is distinctly separated from the tip of the 


1. Tmniotoca lateralis Agassiz is a viviparous blue perch found 
in shallow waters along the entire western coast of North 
America, between San Diego, United States, and Vancouver Is- 
land, Canada. 

66, 1 Blanco: Caudal Fin of the Blue Perch 43 

2. The developmental stages of the homocercal caudal fin of 
the blue perch are here described in detail for the first time. 

3. The significant changes of the embryonic caudal fin of a 
larva 3 mm long to the homocercal caudal fin of the adult 21 
cm long are as follows : 

(a) The presence of an axial lobe in a 7-mm larva represents 
a characteristic feature of its embryonic development. 

(6) The change of the protocercal caudal fin to heterocercal 
caudal fin in a larva 12 mm long, and the appearance of an 
anlage of six asymmetrical cartilaginous hypurals, each having 
one or more caudal rays. 

(c) A larva 14 mm long has the beginnings of the neural 
spines which develop above the upturned notochord. 

(d) The caudal fin of a larva 16 mm long has seven hypurals, 
the first four parallel to each other; the third, fourth, and fifth 
have a common base. 

(e) The axial lobe at this stage is evidently lost. 

(/) In a larva 21 mm long the urostyle is present as the 
result of the shortening and thickening of the posterior part of 
the notochord. Three dorsal caudal radials are at this stage 
inserted between the urostyle and the last neural spine. 

(g) A larva 41 mm long has apparently hypurals which are 
joined to the penultimate vertebra, while the segmented dermo- 
trichia are at this stage well developed. 

(h) The caudal fin of a young perch 51 mm long shows a 
slight difference from that of an adult homocercal caudal fin of 
a specimen 21 cm long. The urostyle is much reduced in the 
adult, and the dorsal caudal rays are distinctly separated from 
the tip of the urostyle. 


Dermotrichia. a caudal fin ray. 

Protocercal. A protocercal caudal fin is a primitive type of a caudal, 
externally and internally symmetrical, indicating that the tail has not 
undergone any modification from the original form. 

HETWaioCERCAL. A heterocercal caudal is one in which an upbending of the 
vertebrae has taken place, and in which the hypaxial and epaxial ele- 
ments and the caudal lobe of the fin are asymmetrically placed. 

Homocercal. A homocercal caudal fin is a specialized heterociercal cau- 
dal, externally symmetrical; the majority of the fin rays are sup- 
ported externally by the hypaxial elements; a urostyle, present in the 
larval stage, persists for a time in the adult. 

Urostyle. The urostyle is an elongated conelike termination of the 
vertebral column, representing the fusion of several last vertebree. 

Hypaxial. Morphological structures ventral to the chordal axis. 

Epaxial. Morphological structures dorsal to the chordal axis. 

44 The Philippine Journal of Science 

Hypural. Any hsrpaxial structure connected directly with the chorda! 

axis to which one or more fin rays are attached. 
Epural. Any epaxial element having direct connection with the chordal 

axis to which one or more fin* rays are attached on its distal end. 
Radial. Synonymous to "interspinous bone*' and "somatid;** a daggerlike 

bone having no direct connection with the vertebral column ; it supports 

one or more fin rays distally. 
Last Vertebral Segment. This term refers to the last centrum and the 

urostyle, if such structures exist; the last centrum is never perfect; 

it is usually conelike with the apex directed posterodorsally. 


Agassiz, a. On the young stages of some osseous fishes; I. Develop- 
ment of the tail. Proc. Am. Acad. Arts and Sci. 13 (1878) 117. 

Eigenmann, Carl H. On the viviparous fishes of the Pacific Coast of 
North America. Bull. U. S. Fish Comm. 12 (1894) 381-478, pis. 93- 

Ryder, J. A. Evolution of the fins of fishes. Report of the Comm. of 
Fish and Fisheries. Wash, (1884) 981--1107. 

Whitehouse, F. H. The caudal fin of fishes; Preliminary paper. Proc. 
Roy. Soc. B. 82 (1910) 134-143. 

Whitehouse, F. H. The caudal fin of the Teleostomi. Proc. Zool. Soc. 
London pt. 2 (1910) 590-626, pis. 47-50, figs. 1-33. 



axl, Axial lobe. fn, Fin fold, 
a, 6, c, df 6, /, g. Cartilaginous hsy Haemal spine. 

hypural areas. hy, Hypural. 

cemt, Centrum. Iv, Last vertebra. 

der, Dermotrichia, ns, Neural spine. 

dcr, Dorsal caudal radial. nt, Notochord. 

ep, Epural. ur, Urostyle. 
/, Permanent caudal fin. 

Plate 1. T^niotoca LATERAiiis Agassiz 

Fig. 1. A larva 3 mm long. 

2. Caudal of a larva 7 mm long. 

3. Caudal of a larva 12 mm long. 

4. Caudal of a larva 14 mm long. 
Figs. 5 and 6. Caudal of a larva 16 mm long. 

Fig. 7. Caudal of a larva 21 mm long. 

8. Caudal of a larva 41 mm long. 

9. Caudal of a young perch 55 mm long. 

10. Adult homocercal caudal of a female 21 cm long. 


Blanco: Caudal Fin of the Blue Perch.] 

[Philip. Journ. Sci., 56, No. 1. 




By F. L. Stevens 

Professor of Plant Pathology of the Unversity of Illinois; Charles Fuller 

Baker Memorial Professor of Plant Pathology (1 930-1 9S1) 

University of the Philippines 



Of the Department of Plant Pathology, College of Agriculture, University 
of the Philippines, Los Banos 


The purpose of the present article is to record the Meliolineae 
collected by the senior author during his stay of eight months 
in the Philippines as Charles Fuller Baker Memorial Professor 
of Plant Pathology (1930-1931) that may be new, or upon hosts 
not previously recorded, or otherwise of especial interest. The 
arrangement is that of the senior author's monograph of the 

Specimens of all fungi noted in this article have been de- 
posited in the herbarium of the University of Illinois, Urbana, 
Illinois ; the herbarium of the Bureau of Science, Manila, Phil- 
ippine Islands; and the herbarium of the College of Agricul- 
ture, Los Baiios, Laguna, Philippine Islands. In addition, when 
material was of a quantity sufficient to warrant it, duplicate 
specimens were placed in several leading mycological centers in 
America. We are much indebted for determinations of hosts 
to Ranger Mamerto Sulit, of the School of Forestry, Los Banos; 
to Dr. Eduardo Quisumbing, botanist of the National Museum 
Division of the Bureau of Science; and to Dr. E. D. Merrill, 
director-in-chief of the New York Botanical Garden. 

* Contribution No. 977 from the experiment station of the College of Agri- 
culture, Los Baiios, Laguna, Philippine Islands, and of the Department of 
Botany of the University of Illinois, Urbana, Illinois. Published with the 
approval of the dean of the College of Agriculture. 

*Ann. Mycol. 25 (1927) 405; 26 (1928) 165. 



The Philippine Journal of Science 
Geims AMAZONIA Tkeissen 


Uf AZONIA PBREGKINA (Sydow) Sydiow. 

Group number S101.4230. 

On Myrsinaceae: Maesa laxa. This species is widely distrib- 
uted in Luzon, collections having been made in Benguet, La 
Union, Ilocos Sur, Bontoc, and Laguna. 

PlO. 1. 

Genus MELIOLINA Sydow 

MELIOLINA SAURAUIAB sp. hot. Fig. 1. •. 

Colony epiphyllous, 3 to 15 mm in diameter, black, diffuse, in- 
definite. Mycelium 6 {i. thick, crooked, smooth, branching at very 
acute angle. Spot none. Capitate hyphopodia and mucronate 
hyphopodia none. 

5«, 1 Stevens and Roldan: PhUippine Meliolinex 49 

iQnf 'fcfn'^^. '^^ ^""^ "^^'^^^"^ '^^ ^^^^- Perithecia globose, 
130 to 190 (X m diameter, very rough, surfaced with conic projec- 
tions about 10 to 30 ^ high. Asci evanescent, 2- to 4-spored 
Spores 4-septate, long and narrow, 57 to 70 by 13 a end cells 

Group number 3100.6220. 

On Dilleniacese : Saurauia latibractea. Naguilian Road, Ben- 
guet, Luzon, January 5, 1931, No. 1480. 

This exceedingly interesting species falls within the section of 
the genus having no setae, which section has only five species 
and three of these are recorded as parasitic upon other Melio- 
linese. Of the two reported on phanerogams, one has 3.septate 
spores, thus leaving only one Meliolina that has the formula 
3100; namely, M. megalospora (Speg.) Stevens, on Jodimi. 
Comparison of the senior author's specimen with the type of 
this shows very distinct differences in colony and mycelial char- 
acter, the present species having a thinner mycelium and a 
larger, looser colony. It is interesting that another Meliolina, 
M. malacensis (Sacc.) Stevens, is described on the Dilleniace» 
from Singapore; this one, however, has long and much-branched 


Group number 2140.5342. 

In the senior author's monograph of the Meliolinese, page 
419, it was suggested that though M. yatesii was originally de- 
scribed by Sydow as being on Viburnum, in reality the host may 
be one of the Myrtacese. Dr. E. Quisumbing has definitely 
determined for us that the host of the type material is not a 
Viburnum but a Eugenia. Therefore, the synonymy suggested 
by one of us in the monograph is confirmed. 

Genus IRENE Theissen and Sydow 


Group number 3201.4320. 

On Dilleniaceae : Saurauia elegans. Three collections were 
made on the road from Baguio to Naguilian, Mountain Province, 

Genus IRENOPSIS Stevens 


Colony 2 to 20 mm in diameter, amphigenous, black, indefi- 
nite. Mycelium 6 to 7 {x thick, black, smooth, slightly crooked. 
Spot none. Capitate hyphopodia alternate, not crowded, usual- 

28909S i 

50 The Philippine Journal of Science 1935 

ly about 30 \l apart, often less close. Stalk cell short, 4 to 6 [i., 
head cell pyriform, regular or slightly irregular. Mucronate 
hyphopodia ampuUiform, alternate or opposite. 

Perithecial setsB arising from the lower half of the peri- 
thecium, 70 to 105 ^ long, 10 jx thick, simple, black, straight 
or slightly curved. Mycelial setae none. Perithecia globose, 
smooth, 100 to 115 [i. in diameter. Asci evanescent. Spores 
4-septate, 40 by 13 to 15 ^. 

Group number 3401.3220. 

On Moracese: Ficus variegata. Naguilian Road, Benguet,. 
Luzon, January 5, 1931, Nos. 1566 (type) and 1528. 

No species of Irenopsis has heretofore been described on Ficus 
or any other genus of the Moracese. This species is very clearly 
characterized by the entire absence of mycelial setae and the 
presence of a few short black setae arising from the base of the 

IRENOPSIS CORONATA (Speff.) SteTens yar. PHILIPPINENSIS rar. nov. Fig, Z, J. 

Colony very minute, almost invisible. Mycelium dark, smooth, 
6 to 7 [X thick. Spot none. Capitate hyphopodia alternate, in 
some colonies few, in others more abundant. Stalk cell short, 
3 to 4 [x, head cell pyriform, regular. Mucronate hyphopodia 
ampuUiform, often very abundant, present almost to the exclusion 
of the capitate hyphopodia. 

Perithecial setae few, 6 to 8, growing from the lower part of 
the perithecium, 100 to 130 \l long, not uncinate. Mycelial 
setae none. Perithecia globose, black, smooth, except for a few 
set^ arising from near the base. Asci evanescent. Spores 4- 

Group number 3401.4220. 

On Tiliaceae: Columbia serratifolia. Agricultural College, 
Laguna Province, Luzon, September 10, 1930, No. 510. 

The irregular arrangement of the hyphopodia is striking. 
Sometimes the capitate hyphopodia are abundant and regularly 
placed; at other times they are very irregularly placed, giving 
the mycelium almost the appearance of a branch without 
hyphopodia. In some colonies the capitate hyphopodia are 
exceedingly rare, while the mucronate hyphopodia are very; 

This variety differs from other varieties of this species in its 
variable ratio between mucronate and capitate hyphopodia and 
in the shape of its capitate hyphopodia, which are pyriform 
while in the other varieties the capitate hyphopodia are shorter. 

56, 1 Stevens and Roldan: Philippine Meliolinese 51 

almost globose. It is striking to note that the only Meliolinese 
described on the Tiliaeese are Irenopsis coronata and three varie- 
ties of this fungus. Two species of Irenina that were originally 
described on other families are said to be found on Tiliaeese, but 
they may rest upon erroneous determinations. Though only this 
one species and its four varieties are known upon the Tiliaeese, 
they are widely distributed, being recorded from Paraguay, 
Africa, Brazil, Argentine, Porto Rico, Santo Domingo, and the 

UtENOPSIS CORONATA (Spear.) Stevens var. TRIUMPETTAE (Stevens) Stevens. 

Group number 3401.4220. 

On Tiliaeese: Triumfetta bartramia. 

Hyphse of considerable length without hyphopodia frequently 
occur, especially within areas near the perithecia. 

A Meliola on Triumfetta which has been reported from the 
Philippines as M. arachnoidea is probably this variety. 

The remarkable agreement in the character of the colony on 
the leaf and the morphology of this fungus on Triumfetta from 
such widely separated parts of the world as Porto Rico, South 
Africa, and the Philippine Islands is noteworthy. 

Genus IRENINA Stevens 

IRENINA ACALYPHAE sp. nov. Fiar. 1, c 

Colonies small, mostly less than 2 mm in diameter, amphige- 
nous, densely black, circular. Mycelium 7 to 8 (x thick, densely 
crowded. Spot none. Capitate hyphopodia opposite, densely 
crowded, touching. Stalk cell short, 3 to 4 (x, head cell subglo- 
bose to subcubical. Mucronate hyphopodia ampulliform. 

Perithecial setse none. Mycelial setse none. Perithecia glo- 
bose, 130 to 145 [i. in diameter, very rough with conic protuber- 
ances up to 30 [JL high, though with no true larviform appendages. 
Asci evanescent. Spores 4-septate, 47 by 16 [x. 

Group number 3102.4220. 

On Euphorbiacese: Acalypha. Acop's, Benguet, Luzon, De- 
cember 30, 1930, No. 1211. 

Although five species of Irenina are recorded as upon the Eu- 
phorbiacese, none of them has opposite capitate hyphopodia. 
The extreme roughness of the perithecial wall suggests the ge- 
nus Irene, but no actual vermiform appendages were found, nor 
is any species of Irene with opposite hyphopodia known on this 
host family. The formula is close to that of Amazonia acaly- 
phae, but the perithecia here are free and globose and show no 

52 The Philippine Journal of Science i5>35 

kinship to Amazonia. The chief characteristic of the species is 
the very rough perithecium and the very dense character of the 

IRENINA RUBI sp. nov. Fig, h d. 

Colony very small, nearly invisible, usually consisting of only a 
few mycelial strands. Mycelium 6 to 7 [x thick. Spot none. 
Capitate hyphopodia alternate, usually about 23 (x apart. Stalk 
cell short, 3 to 4 [x, head ceU globose, regular. Mucronate hy- 
phopodia ampuUiform, 16 to 20 \l long. 

Perithecial setse none. Mycelial setae none. Perithecia glo- 
bose, smooth, 130 to 145 [jl in diameter. Asci evanescent. Spores 
4-septate, 33 to 36 by 14 \l. 

Group number 3101.3220. 

On Rosacese: Rubus rosaefolius. Naguilian Road, Benguet, 
Luzon, January 6, 1931, No. 1549 (type) ; Sariaya, Tayabas 
Province, Luzon, August 9, 1930, No. 193. 

On Rnhus molncanv^. Mount Santo Tomas, Benguet, Luzon, 
December 31, 1930, No. 1361. 

IRENINA RUBI var. ANGULATA rar. hot. Fig. 1. c. 

This variety is like the species except that the capitate hypho- 
podia have head cells that are irregular to angular and large, to 
17 (1. across. 

On Rosacese: Rubus molucanus. Naguilian Road, Benguet, 
Luzon, January 6, 1931, No. 1461 (type). 

Rubus rosaefolius, Naguilian Road, Benguet, Luzon, January 
6, 1931, No. 1472. 

These two forms of Meliokt, each occurring on two species of 
Rubus in the same region of Luzon, are interesting. Sometimes 
either one or the other of the species is found alone upon a 
given collection. On other collections both varieties may 
occur side by side upon the same leaf. There is, however, no 
evidence of intergrading of the two forms; one has the small 
regular head cells, the other the large irregular ones. 

The Meliolinese upon Rosacese in the Philippines are especial- 
ly interesting since of the five species previously described three 
had 3-septate spores and two had 4-septate spores. Moreover, 
the 3-septate species were of wide distribution, including the 
United States, South America, Japan, Porto Rico, Hawaii, and 
Africa, while the two 4-septate species were known only from 
the Argentine and Brazil. It appears, therefore, that the pre- 
ponderant type upon Rosacese was that of 3-septa. 

»«'! Stevens and Roldan: Philippine Meliolinese 53 

In the Philippines no species had heretofore been reported 
upon Rubus, doubtless due to their very minute, inconspicuous 
colonies. The species now found upon Rubiis are all of the 4- 
septate type, none with 3-septate spores. 

It may be possible that the Philippine Meliolinese upon Rubus 
have comparatively recently acquired tenancy upon that host, 
coming perhaps from some other host, and that the present 
poorly developed colony may be due to the fact that they have 
not become as well adapted to their host as have the 3-septate 
forms, which have had much longer residence upon Rosacese. 

IRENINA ANGUSTISPORA sp. nov. Fig. 1. f. 

Colony 1 to 3 mm in diameter, thin, hypophyllous. Mycelium 
4 to 5 [JL thick, very crooked, often zigzag, smooth. Spot none. 
Capitate hyphopodia alternate or unilateral, usually from 13 to 
35 [Ji apart. Stalk cell short, 3 to 4 [x, head cell subglobose, oblong 
or cuneiform. Mucronate hyphopodia ampuUiform. 

Perithecial setae none. Mycelial setae none. Perithecia glo- 
bose, smooth, 146 to 175 [x, black, very rough with conic black 
projections about 16 ijl high, which simulate but are not larviform 
appendages. Asci evanescent. Spores 4-septate, 33 by 7 to 9 (x. 

Group number 3101.2120. 

On Rubiaceae : Neonauclea sp. Naguilian Road, Benguet, Lu- 
zon, January 7, 1931, No. 1620. 

This species is chiefly characterized by its very rough perithe- 
cium, very crooked mycelium, and very narrow ascospores. It 
differs from all Ireninse on the Rubiaceae in these regards. 

IRENINA ANGUSTISPORA Stcveng var. LAEVIS Tar. nov. Figr. 1, g. 

On Rubiaceae: Neonauclea. Kennon Road, Benguet, Luzon, 
January 8, 1931 ; No. 1633. 

This variety is like the species except that the perithecia are 


Colony epiphyllous, of indefinite shape, not very dark, myce- 
lium smooth, crooked, about 5 p. thick, anastomosing profusely, 
forming a network. Spot none. Capitate hyphopodia alternate, 
usually 33 to 50 [a apart. Stalk cell short, 4 to 11 jx long, head 
cell ovate to pyrif orm, regular. Mucronate hyphopodia ampuUi- 
form, pale, thin. 

Perithecial setae none. Mycelial setae none. Perithecia glo- 
bose, 150 ^ in diameter, rough due to conical surface cells. Asci 
evanescent. Spores 4-septate, 36 by 17 {x. 

Group number 3101.3220. 

54 The Philippine Journal of Science v^zh 

On Verbenacese: Callicarpa magna. Naguilian Road, Ben- 
guet, Luzon, January 6, 1931, No. 1468 (type) ; Santo Tomas, 
La Union Province, Luzon, December 31, 1930, No. 1291. 

Of the five species of Ireninxi recorded as upon members 
of the Verbenacese and that have alternate capitate hyphopodia, 
only /. glabroides and /. vitis have spores as small as those 
in the present species, and each of these differs markedly in 
mycelial character, hot anastomosing freely as is the case with 
the present species. The distinctive character of this species 
is the abundant anastomosis of the mycelium, making the colony 
into a distinct network. 


Colony 1 to 4 mm in diameter, loose, epiphyllous. Mycelium 
6 jx thick, crooked. Spot none. Capitate hyphopodia alternate, 
usually 16 to 26 \l apart, sometimes 70 [jl. Stalk cell to 10 [i. 
long, head cell large, angular, 16 [x. Mucronate hyphopodia am- 
puUiform, scarce. 

Perithecial setse none. Mycelial setse none. Perithecia glo- 
bose, 145 to 160 jx, rough with conic protuberances, about 15 \k 
high. Asci evanescent. Spores 4-septate, 43 by 16 [jl. 

Group number 3101.4220. 

On Acanthacese: Thunbergia alata. Kennon Road, Benguet, 
Luzon, January 8, 1931, No. 1642. 

This in its larger, more irregular head cells, and its much less 
dense and larger colony shows no resemblance to /. irregularis. 


Group number 3101.2220. 

On Meliacese: Sandoricum koetjape. Agricultural College, 
Laguna Province, and Quezon Forest Park, Tayabas Province, 

IRENINA BINUOSA sp. nor. Fig. 1, I. 

Colony hypophyllous, thin, scattered, indefinite, to 2 cm in 
diameter. Mycelium very crooked, so as to give the appearance 
of anastomosing, thin, 3 to 4 y.. Spot none. Capitate hyphopo- 
dia alternate or unilateral, usually about 25 [x apart. Stalk cell 
short, 3 to 4 (i., standing out at right angles to the mycelium, 
head cell cylindrical or ovate. Mucronate hyphopodia ampulli- 

Perithecial setse none. Mycelial setse none. Perithecia 
globose, black, somewhat rough, 140 to 175 \l in diameter. Asci 
evanescent. Spores 4-septate, 50 to 57 by 17 ^. 

Group number 3101.4220. 

56,1 Stevens and Roldan: Philippine MeliolinesB 55 

On Euphorbiaceae: Glochidion sp. Balete Pass, Nueva Ecija 
Province, Luzon, January 9, 1931, No. 1744 (type) . 

This species does not have capitate hyphopodia close or glo- 
bose as in M. alchoreae, nor very few as in /. subapoda and 
/. verrucosa. It differs from /. alchomeae in its crooked my- 

IRENINA SUBAPODA (Sydow) Stevens. 

Group number 3101.3220. 

On Euphorbiaceae: Mallotus philippinensis. Balete Pass, 
Nueva Ecija Province, Luzon. 

IRENINA UNCARI^ (Rehm) Stevcn». 

Group number 3102.2220. 

On Rubiaceae: Uncaria perrottetiL Collections were made 
from Laguna, Benguet, and Tayabas Province, Luzon. 


Group number 3101.4220. 

On Combretaceae : Quisqualis indica. Santa Cruz, Laguna 
Province, Luzon. 

None of the Meliolineae has been reported before upon Quis-- 

Genus MELIOLA Fries 

Group 1. Spores 2- or 3-septate ; formula V2 - — . or 2 ^. or % . 

MELIOLA PALAQUII ep. nov. Fi^. 2, a. 

Colony 4 to 6 mm in diameter, often coalescing, black, cir- 
cular, amphigenous but mainly hypophyllous. Mycelium 6 pt 
thick, slightly crooked, black. Spot blanched. Capitate hypho- 
podia alternate, not crowded, 16 to 135 {jl apart. Stalk cell long, 
to 13 (JL, head cell large, to 23 (ji, very irregular and angled. Mu- 
cronate hyphopodia ampulliform, thick, long. 

Perithecial setae none. Mycelial setae about 200 \i» long, 5 to 
6 [x thick at base, tapering to 3 |a thick at apex, obtuse, mastigal. 
Perithecia sparse, globose, smooth, 145 to 175 \l in diameter. 
Asci evanescent. Spores 3-septate, 50 to 54 by 16 to 20 |jl, the 
two central cells large, the end cells smaller and somewhat 

Group number 211.5221. 

On Sapotacese: Palaquium sp. Mount Maquiling, Laguna 
Province, Luzon, January 18, 1931, No. 1900. 

This beautiful species differs markedly in the very irregular 
head cells from all other species found upon the Sapotaceae, 
also in having 3-septate spores and the long whiplike setae. It 
is the only species known on this family with 3-septate spores. 


The Philippine Journal of Science 
Group 3. Formula 314~. or 311. 


IIELIOLA GANOPHYLLI sp. noy. Fiff. 3, k. 

Colony 1 to 3 mm in diameter, very black, very dense. My- 
celium 6 to 7 ix thick, strands growing side by side so as to 
cover completely the subtending surface. Spot none. Capitate 
hyphopodia alternate, closely crowded. Stalk cell short, 3 to 
4 (A, head cell globose. Mucronate hyphopodia ampulliform. 

Fig. 2. 

Perithecial setae none. Mycelial setae numerous, 190 to 300 |a 
long, 7 ^ thick at base, apex bearing two, three, or four short, 2- 
to 6-[x teeth, or rarely twice branched, and primary branches 
20 }^ long. Perithecia globose, smooth, 145 to 160 [x in diameter. 
Asci evanescent. Spores 4-septate, 47 to 50 by 20 [a. 

56,1 Stevens and Roldan: Philippine Meliolinese 57 

Group number 311-1.4221. 

On Sapindacese : Ganophyllum falcatum. Kennon Road, 
Benguet, Luzon, January 8, 1931, No. 1671. 

This species differs from M. sapindacearum in having no op- 
posite capitate hyphopodia; from M. sapindi, which has larger 
and more irregular head cells and a less dense colony ; it is close 
to M. paulliniae var. dentata but differs from it in head cells and 
in density of colony. No Meliola has heretofore been recorded 
upon this host genus. 


Group number 3141.4231. 

Collection was made at Balete Pass, Nueva Ecija Province, 
Luzon, on Murraya paniculata. It had not been before reported 
on this genus. 


Group number 311-1.3223. 

This species was described by Sydow as on an unknown host. 
The host of the type specimen was determined for us by Dr. 
E. Quisumbing as Dracontomelum dao, of the Anacardiacese. 

MELIOLA MICROMELI sp. noT. Wig. 2, b. 

Colony 1 to 3 mm in diameter, amphigenous, more often epi- 
phyllous, black, very dense, crustose. Mycelium 6 \l thick, 
smooth, crooked. Spot pale, showing from both sides of the 
leaf. Capitate hyphopodia usually alternate, rarely opposite, 
densely crowded, usually less than 6 \l apart. Stalk cell short, 
3 to 4 |x, head cell ovate, regular. Mucronate hyphopodia am- 
puUiform, narrow. 

Perithecial setse none. Mycelial setse very numerous, 220 
\k long, 7 PL thick at base, dentate with two to five short teeth, 
usually 3 to 5 [JL long, sometimes forked with two branches 20 
[A long and these with short teeth. Perithecia globose, smooth, 
145 to 150 [X in diameter. Asci evanescent, 2-spored. Spores 4- 
septate, 47 to 53 by 16 to 21 \l. 

Group number 31i-3.5221. 

A strikingly unique character of this fungus is that the 
mycelial strands near the edge of the colony for a zone some 
150 to 200 [x wide, run parallel with each other and touching 
each other and in this zone they are almost or quite devoid of 

On Rutaceae : Micromelum minutum, San Jose to Balete Pass, 
Nueva Ecija Province, Luzon, January 6, 1931, No. 1726. 

58 'I'^^ Philippine Journal of Science 1935 

This species differs from all of the nineteen species of Meliola 
described as on the Rutacese. It is most nearly related perhaps 
to M. tenella, M. hamhusae var. atlantiae, and M. evodeae. From 
M. tenella and M. hamhusae var. atlantiae it differs in having a 
dense colony. Meliola evodiae does have a dense colony but does 
not have the character of colony given above, nor are its setse 
so long. 

Group 4. Formula 313-. 

BCELIOLA BENGUETENSIS sp. xiov. Fig. 2, c. 

Colony 1 to 6 mm in diameter, amphigenous, very black, 
very dense. Mycelium 6 \l thick, smooth. The mycelium at the 
edge of a colony constituting a zone about 200 mm wide that 
is pale and devoid of hyphopodia and setse. Spot pale yellow, 
devoid of chlorophyll, visible from both sides of the leaf. Ca- 
pitate hyphopodia alternate or unilateral, somewhat crowded. 
Stalk cell short, 3 to 4 (x, head cell subglobose, regular. Mucro- 
nate hyphopodia ampullif orm, very irregular, numerous on some 
hyphse, mostly opposite. 

Perithecial setse none. Mycelial setse very black, 250 to 270 
\h long, 10 \K thick at base, with 2, 3, or 4 short, 6- to 7-[j(. teeth 
at apex. Perithecia globose, smooth, 145 to 160 \l in diameter. 
Asci evanescent. Spores 4-septate, 47 to 50 by 17 \i.. 

Group number 3131.4221. 

On Sapindacese : Otophora sp. Kennon Road, Benguet, Luzon, 
January 8, 1931, No. 670 (type) ; San Jose-Balete Road, Nueva 
Ecija Province, Luzon, January 10, 1931, No. 1794. 

The only species of Meliola upon the Sapindacese of the for- 
mula 3131. are M. paulliniae var. dentata and M. sapindi, the 
latter differing from the present form in having irregular head 
cells, while the former is distinguished by its thin colony. 

The colony edge of the species is characteristic. 

BI:ELI0LA PISTACIAE sp. nov. ¥ig. 2, d. 

Colony 1 to 3 mm in diameter, circular, densely black, amphi- 
genous but mainly epiphyllous, also caulicolous. Mycelium 6 [x 
thick, densely crowded. An outer zone of each colony consists 
of paler mycelium that bears no hyphopodia or setse. Spot 
none. Capitate hyphopodia alternate, very rarely opposite, close- 
ly crowded. Stalk cell short, 3 {/., head cell subglobose, regular. 
Mucronate hyphopodia ampullif orm. 

Perithecial setae none. Mycelial setae 233 to 277 jx long, 
8 p. thick at base, apex bearing two or three short, 5 to 9 [x, 

56, 1 Stevens and Roldan: Philippine Meliolinese 59 

acute teeth. Perithecia globose, smooth, 145 ix in diameter. 
Asci evanescent. Spores 4-septate, 43 by 16 (i.. 

Group number 3131.5221. 

On Anacardiaceae : Pistacia sp. San Jose to Balete Pass, 
Nueva Ecija Province, Luzon, January 9, 1931, No. 1712. 

Three species of Meliola of the formula 3131. have their type 
forms on the Anacardiaceae. Two of these have cristate setal 
tips, thus disagreeing with the present species. From the third 
species, M. brachyodonta Syd., it differs decidedly in the colony 
character described above. 

MELIOLA THEMEDAE sp. nor. Fig. 2, e. 

Colony amphigenous, 1 to 3 mm in diameter, black, dense, al- 
most crustose. Mycelium 6 to 7 ix thick, straight, longitudinally 
on the leaf, crooked transversely. Spot none. Capitate hypho- 
podia alternate. Stalk cell to 10 [x long, head cell irregular, an- 
gular. Mucronate hyphopodia ampulliform, small. 

Perithecial setse none. Mycelial setae stiff, straight, black, 
160 to 190 [JL long, 10 ix thick at base, somewhat enlarged and 
divided into numerous short teeth at tip, subcristate. Perithecia 
globose, smooth, 130 to 145 [x. Asci and spores not seen. 

Group number ?131.??21. 

On Gramineae: Themeda gigantea. Mufiioz, Nueva Ecija 
Province, Luzon, October 3, 1930, No. 794. 

The setal tips are quite striking, different from any others 
that the senior author has seen. 

Group 5. Formula 81i~. or SU-. or 31i-. 


Group number 31|~1.3221. 

This is a well-defined species characterized by its irregular 
setal tips on setse that abound only near the perithecia. Occa- 
sionally a leaf bearing M. desmodii may also have colonies of 
M. heterocephala. The two colonies on close observation can be 
distinguished by their habit and by the capitate hyphopodia. 
On Desmodium occur several forms of Meliola recorded as dis- 
tinct species, and the number on the Leguminosae is very large. 
Many of these show a common character, probably indicating 
a common ancestry, in their small, globose, capitate hyphopodia. 
These segregate into species since some have the hyphopodia 
strictly opposite, others strictly alternate, etc.; also by varia- 
tion in the length of the mycelial setse or in possessing charac- 
teristic setal tips. Frequently two nearly related species occur 

60 The Philippine Journal of Science isss 

on the same leaf and at first may then be puzzling to the tax- 


Group number 31i-1.4223. 

This species is very common in the Philippines and is much 
more variable than the description of Sydow would lead one to 
expect. Sydow's description gives the setse as 350 to 700 ix 
long, while on some of our specimens we find them over 1,400 
p., others being as short as is indicated by Sydow's description. 
In length of setse this species resembles M. magna Stevens de- 
scribed from Costa Rica on Nectandra. No other species upon 
the LauracesB and, indeed, but few other species of Meliola have 
setse more than 1 mm long. This species differs strikingly from 
M. magna in its smaller spores. 

MELIOLA NEPHELIICOLA «p. nov. Fig. 2, f. 

Colony 3 to 10 mm in diameter, hypophyllous, black, indefinite, 
adhering very closely to the leaf surface. Mycelium 4 to 5 ii 
thick, crooked. Spot none. Capitate hyphopodia alternate or 
opposite. Stalk cell short, 3 to 4 {x, head cell subglobose. Mu- 
cronate hyphopodia ampulliform. 

Perithecial setse none. Mycelial setse 185 to 470 \k long, 10 
fi thick at base, either simple and acute or with a few very 
minute teeth. Perithecia globose, smooth, 150 to 190 \l in diam- 
eter. Asci evanescent. Spores 4-septate, 40 by 13 \l. 

Group number 31^-3.3222. 

On Sapindacese: Nephelium intermedium. Mount Maqui- 
ling, Laguna Province, Luzon, July 22, 1930, No. 77. 

This differs from M. capensis in that it has a much less dense 
colony; its hyphopodia are not always opposite as they are in 
M. capensis. It differs from M. variaseta in its acute setal 
apices. It differs from M. nephelii in shape of hyphopodia and 
abundance of setse. 

Group 6. Formula 312-. or 31i~. 


Group number 3121.5343. 

This species was described by Sydow as on an unknown host. 
Ranger Mamerto Sulit, of the School of Forestry, Los Banos, has 
kindly determined the host of the type specimen as Eugenia, 
of the Myrtaceae. It appears to be closely related to M. dens'a 
but differs from it in having much shorter setae. 

56,1 Stevens and Roldan: Philippine Meliolinese 61 

MELIOLA ANDROPOGONIS sp. nor. Fiir. 3, 1. 

Colony epiphyllous, circular or ovate, 1 to 3 mm in diameter, 
black, dense. Mycelium 6 to 7 (jl thick, straight. Spot none. 
Capitate hyphopodia alternate. Stalk cell short, 3 to 4 jx, head 
cell ovate, subglobose or more often angular. Mucronate hypho- 
podia ampulliform. 

Perithecial setse none. Mycelial setae simple, obtuse, hamate, 
140 to 220 [A long. Perithecia globose, smooth. Asci evanes- 
cent. Spores 4-septate. 

Group number 3121. 1. 

On Gramineae: Andropogon halepensis. Naguilian Road, 
Benguet, Luzon, January 7, 1931, No. 1577. 

Of the thirteen species of Meliola recorded on the Graminese 
only M, panici has been reported as on Andropogon. The ham- 
ate setae and the irregular hyphopodia characterize the present 

Group 7. Formula 3113. 

MELIOLA STMPHOREMAE sp. nov. Fiff. 2. g. 

Colony 5 to 10 mm in diameter, indefinite, black, hypophyl- 
lous. Mycelium 6 [x thick, somewhat crooked. Spot none. 
Capitate hyphopodia opposite or alternate, mainly opposite, pairs 
somewhat crowded, often only 13 ^ apart. Stalk cell short, 3 
to 4 [X, head cell irregularly ovate or oblong, 10 to 13 \i. long. 
Mucronate hyphopodia ampuUiform, very numerous, alternate 
or opposite, mostly opposite, 30 (x long, 7 [a wide. 

Perithecial setae none. Mycelial setae numerous, 300 to 585 ix 
long, 10 [A thick at base, black, simple, obtuse or acute. Peri- 
thecia globose, smooth, 190 [jl in diameter. Asci evanescent. 
Spores 4-septate, 40 by 17 {x. 

Group number 3113.3223. 

On Verbenaceae: Symphorema luzonicus. Mount Maquiling, 
Laguna Province, Luzon, No. 655. 

Only one species of the formula 3113, M. callicarpae, is re- 
corded as on the Verbenaceae and that has shorter setae and 
smaller, more regular head cells and is a very distinct species. 
On the type material there is another Meliola, undetermined, 
that has strictly alternate capitate hyphopodia and the colony 
of which is thinner, more diffuse. Both of these species are 
parasitized by a Helminthosporium, which, however, is of much 
more dense growth on the undetermined Meliola than upon M. 
symphoremae, indicating that these parasites upon the MelioUi 
are biologically differentiated as to their hosts. 

62 The Philippine Journal of Science 1935 


Group number 3113.4231. 

The specimens on Acacia confusa in the Philippine Islands 
show no differences from the type material collected by the se- 
nior author on Acacia koa in Hawaii. Though these two species 
of Acacia are both phyllodinous, they are not closely allied. It 
seems probable that both species of hosts derived the Meliola 
from some remote ancestor, probably Australian. 


Group number 3113.4222. 

A collection was made on Mount Maquiling on Cissampelos on 
which genus it has not before been reported. 


Meliola glabrivscula Spegazzini, Rev. Mus. La Plata 15 (1908) 15. 

Group number 3113.3231. 

This species is recorded in the senior author's monograph, 
based on the original description, as of the formula 3112. Study 
of Spegazzini's original specimen shows the hyphopodia often 
to be alternate; therefore, the above formula is correct. 


Group number 3113.3222. 

The form of this species that occurs upon Callicarpa ereoclona 
differs slightly from the type, but this is not other than might be 
expected, owing to the fact that C. ereoclona is densely covered 
with hairs while C. cana is not. 

Group 8. Forirmla 3112. 


Group number 3112.3222. 

On Aglaia dijfusa. Mount Maquiling, Laguna Province, Lu- 
zon. This fungus had not been reported on this genus. 

IIELIOLA PISONIAE sp. nov. Fiff. 2» h. 

Colony 1 to 3 mm in diameter, hypophyllous, circular, densely 
black, velvety, falling away from the leaf when dry. Myce- 
lium 6 to 7 [i. thick, dark, smooth, straight. Spot none. Cap- 
itate hyphopodia opposite, densely crowded, touching. Stalk cell 
short, head cell entire, globose. Mucronate hyphopodia ampul- 
liform, not numerous. 

Perithecial setse none. Mycelial setae straight, stiff, simple, 
acute, black, to 10 [jl thick at base, 540 to 730 ix long. Peri- 

56,1 Stevens and Roldan: Philippine Meliolinese 63 

thecia numerous, globose, 146 to 189 ji, surface slightly rough. 
Asci evanescent. Spores 44 to 50 by 13 to 14 ]i. 

Group number 3112.4223. 

On Nyctaginacese : Pisonia umbellifera. Mount Maquiling, 
Los Baiios, Laguna Province, Luzon, August 8, 1930, No. 373 
(type) . 

Only two species have heretofore been reported as upon spe- 
cies of the Nyctaginacese — namely, M. pulchella and M. erio- 
phora — ^both of which differ markedly from the present species. 
The most striking character of this species is the densely black 
colony composed of mycelium that branches so closely as to form 
a dense mat over the occupied area. The very crowded condition 
of the capitate hyphopodia also adds to the density of the colony. 


Group number 3112.3231. 

Collections were made on Hewittia sublobata, Hewittia bicolor, 
Lepistemon bebictariferum, and an undertermined Ipomoea. 
Though not reported from the Philippines as this species before, 
it has been reported under its various synonyms. 

MELIOLA RUBX sp. nav. Fiff. 2. 1. 

Colony 1 to 3 mm in diameter but very abundant and often 
coalescing to cover large areas of the leaf, black hypophyllous. 
Mycelium 6 to 7 [x thick, slightly crooked, with a strong ten- 
dency to branch at right angles and opposite. Spot none. Cap- 
itate hyphopodia nearly always opposite, usually about 17 {x 
apart. Stalk cell short, 3 to 4 jx, head cell oblong, about 14 by 
7 (x, sometimes somewhat irregular. Mucronate hyphopodia 
long, conic. 

Perithecial setse none. Mycelial setae simple, straight, acute, 
277 [x long, 6 to 7 [X thick at base, quite numerous and distributed 
generally over the mycelium ; that is, not limited to the perithe- 
cial region. Perithecia globose, smooth, about 90 [x in diameter. 
Asci evanescent. Spores 4-septate, 47 to 50 by 17 ix. 

Group number 3112.4211. 

On Rosacese: Rubus molmanus. Naguilian Road, Benguet, 
Luzon, January 6, 1931, No. 1469. 

Comparison of this species with the type of M. glabriuscula, 
the only other Meliola recorded as on the Rosacese, shows very 
distinct differences, particularly as the mycelial setse in M. 
glabriusculd are limited to the neighborhood of the perithecia. 

64 The Philippine Journal of Science 1935 

while in the present species they are widely distributed on the 

MELIOLA AGELAEAE sp. nor. Wig, $, g. 

Colonies amphigenous, irregularly circular, 3 to 8 mm in 
diameter, dense to crustose, black, often variously parasitized. 
Mycelium 6 to 7 [j. thick, dark. Capitate hyphopodia opposite, 
close together. Stalk cell short, 3 to 4 [i, head cell cylindrical, 
oblong. Mucronate hyphopodia ampulliform. 

Perithecial setae none. Mycelial setae simple, acute, 365 \l 
long, 10 to 11 [jt. thick at base. Perithecia globose, smooth, 
140 to 180 {A in diameter. Asci evanescent. Spores 4-septate, 
47 by 17 (X. 

Group number 3112.4222. 

On Connaraceae: Agelaea sp. Quezon Forest Park, Tayabas 
Province, Luzon, November 31, 1930, No. 439. 

The strictly opposite capitate hyphopodia are characteristic. 
Only two species of Meliola, and these with alternate hyphopodia, 
are recorded on the Connaraceae. 

Group 9. Formula 3111. s. obtuse. 


Meliola jasminicola was originally described by P. Hennings 
in 1895 on specimens from Hanoi, Tonkin. In this description 
he made no reference to the position of the capitate hyphopodia, 
though he did state that M. jasminicola was like and related to 
M. polytricha Kalchbrenner and Cooke, which later species has 
alternate hyphopodia. Examination of the type specimen of 
M, jasminicola shows that specimen to have alternate capitate 
hyphopodia. For these reasons, in the senior author's mono- 
graph, page 257, he gave the formula for this species as 3111., 
indicating alternate hjrphopodia. 

Bal states that the capitate hyphopodia are either opposite or 
alternate, a condition that we find in all of our specimens. The 
formula, therefore, should be 3113. 

Mr. Willard H. Watts, who examined many collections of the 
senior author's Philippine material, states : "The young colonies 
appear to have just as many alternate and opposite hyphopodia 
as the older colonies. In many cases the hyphopodia were found 
arising only on one side of the mycelium (fig. 1) . Others showed 
alternate and opposite hyphopodia arising on the same mycelial 
branch (fig. 2). No definite conclusions could be reached as to 
why the hyphopodia should be alternate in some parts and op- 
posite in other parts of the same colony.'* 

56, 1 Stevens and Roldan: Philippine Meliolinese 65 


Group number 3111.4221. 

This was described by Sydow as on an unknown host. Ran- 
ger Sulit, of the School of Forestry, Los Banos, has determined 
the host of the type specimen for the senior author as Pahudia 
rhomboidea, one of the Leguminosse. 

The senior author is unable to identify M. aliena with any 
of the seventeen species of Meliola of the formula 3111. recorded 
as on the Leguminosse. This, therefore, must be regarded as 
the eighteenth of that group. 


Group number 3111.4233. 

This was originally reported by Yates as on an unknown 
host. Ranger Sulit, of the School of Forestry, says that the host 
is Lepisanthes of the Sapindaceae. There are four species of 
similar formula upon the Sapindaceae, but none seems to be iden- 
tified with M. Samarensis which, therefore, stands as on the 


Group number 3111.3221. 

A collection was made in Quezon Forest Park, Tayabas Prov- 
ince, Luzon, on Neonauclea which the senior author believes 
referable to this species. This is the first record of M. psycho- 
triae in the Philippines and the first record of any Meliola on a 


Colonies amphigenous, indefinite, black. Mycelium irregular, 
smooth, black. Spot none. Capitate hyphopodia - alternate, 
close together. Stalk cell short, 3 to 4 [x, head cell ovate to 
pyriform, regular. Mucronate hyphopodia ampulliform. 

Perithecial setae none. Mycelial setae simple, obtuse, straight, 
250 to 360 [JL long. Perithecia globose, smooth, 130 |x in diameter. 
Asci evanescent. Spores 4-septate, 26 to 30 by 10 to 11 (x. 

Group number 3111.2222. 

On Gesneriacese : Epithema sp. Naguilian Road, Benguet, 
Luzon, January 7, 1931, No. 1594. 

Only two species of the formula 3111. with obtuse setae have 
been noted on the Gesneriacese, and both of these have shorter 
setae than the present species. 


Meliola petiolaris Petrak, Anal. Mycol. 29 (1931) 185 non M. petio- 
laris Doidge, 1920. 


66 rAc Philippine Journal of Science 1535 

Group number 3111.5432. 

On Meliacese: Dysoxylum cumingianum. Type locality, Phil- 

MELIOLA ILLIGERAE sp. nov. Fig, Z, a. 

Colony 1 to 3 mm in diameter, epiphyllous, black, velvety. 
Mycelium 8 to 9 ji. thick, crooked. Spot none. Capitate hypho- 
podia alternate, not crowded, sometimes very distant, 170 fx. 
Stalk cell long, 3 to 16 \l. Head cell oblong to clavate, regular 
or slightly irregular, 30 by 17 \k. Mucronate hyphopodia am- 
puUiform, small, 16 to 20 ^ long. 

Perithecial setae none. Mycelial setse 500 to 730 ^ long, 10 
jx thick at base, simple, obtuse, slightly curved. Perithecia 
glol)ose, smooth, 160 to 175 ^ in diameter. Asci evanescent. 
Spores 4-septate, 40 to 43 by 13 to 14 [jl. 

Group number 3111.3223. 

On Hernandiaceae : Illigera luzonensis. Naguilian Road, Ben- 
guet, Luzon, January 5, 1931, No. 1524. 

No Meliola has heretofore been reported upon a plant of this 

MELIOLA MELIACEARUM ep. noT. Fiff. 3» b. 

Colony very thin, pale, hypophyllous, 1 to 2 cm in diameter. 
Mycelium crooked, black, smooth, 6 to 7 pt. thick. Spot none. 
Capitate hyphopodia alternate, scattered, far apart, often 170 {x. 
Stalk cell long, to 10 [a. Head cell oblong to irregular pyriform 
and angled. Mucronate hyphopodia ampuUiform, long, to 33 \l, 
crooked, narrow. 

Perithecial setae none. Mycelial setae long, to 880 ja, flex- 
uous, simple, obtuse, very numerous near the perithecia, rare 
elsewhere. Perithecia globose, smooth, 160 [x in diameter. Asci 
evanescent. Spores 4-septate, 40 by 16 [/.. 

Group number 3111.3223. 

On Meliaceae: Dysoxylum cumingianum. Mount Maquiling, 
Laguna, Luzon, October 7, 1930, No. 824 (type) ; San Jose to 
Balete Pass, Nueva Ecija Province, Luzon, January 10, 1931, 
No. 1767. 

The species is characterized by the very loose colony, the dis- 
tant capitate hyphopodia, the long ampuUiform hyphopodia, and 
the long setae near the perithecia, which distinguish it from 
the numerous species of Meliola described as on the Meliaceae. 
It differs from M. hanaJiaoensis, described on Dysoxylum, in its 


Stevens and Roldan: Philippine Meliolinese 


Fig. 8. 

less dark, less dense colony and in setal tips. It is sparsely rep- 
resented on all specimens collected and is usually accompanied 
by one or more other species of Meliola that, however, were not 
in determinable condition. 

68 The Philippine Journal of Science 1935 


Group number 3111.3221. 

Collections were made upon Premna subscandens and Sympho- 
rema luzonicum. The latter is a new host genus, and the former 
a new Philippine record. 


Colony thin, almost invisible, hypophyllous, indefinite, 3 to 5 
mm in diameter. Mycelium 6 to 7 pi. thick, slightly crooked. 
Spot none. Capitate hyphopodia alternate, not crowded. Stalk 
cell short, 3 to 4 [x, head cell subglobose to ovoid. Mucronate 
hyphopodia ampulliform, straight or crooked. 

Perithecial setae none. Mycelial set^ very few, 140 to 250 
|i long, obtuse. Perithecia globose, black, smooth, 70 to 80 (a 
in diameter. Asci evanescent. Spores 4-septate, 40 to 43 by 
16 \i. 

Group number 3111.4211. 

On Burseracese: Garuga abilo. Muiioz, Nueva Ecija Prov- 
ince, Luzon, October 3, 1930, No. 781. 

This differs markedly in mycelium, hyphopodia, and setae from 
the only Meliola of the formula 3111. so far reported upon the 
Burseracese. No Meliola has heretofore been reported upon Ga- 

MELIOLA MICROSPORA Patoaillard and Gaillard. 

Group number 3111.2121. 

Collection was made on Mount Santo Tomas, Benguet, Luzon, 
on Leucas, on which it has not before been reported. 


Group number 3111.3223. 

The length of setae in various collections varies considerably. 
As described, they are from 400 to 600 \l long. Most of the col- 
lections from the Philippines on Rottboellia fall within these 
limits, though collection No. 1544 had very short setae, barely 
204 ^ long. 

On Gramineae: Miscanthus. Naguilian Road, Benguet, Lu- 

This is the first record of a Meliola on Miscanthus. This is 
probably M. panici, but all the specimens are so heavily over- 
grown by parasites that definite determination is impossible. 


Colony 3 to 10 mm in diameter, epiphyllous, very thin, 
indefinite. Mycelium 6 [x thick, nearly straight, wine-colored, 
smooth. Spot none. Capitate hyphopodia alternate, not crowd- 

66, 1 Stevens and Roldan: Philippine Meliolinese 69 

ed, 30 to 85 pi apart. Stalk cell short, 3 to 4 [jt., head cell clavate, 
to 23 (1. long. Mucronate hyphopodia ampulliform, small, about 
16 [JL long. 

Perithecial setae none. Mycelial setae simple, to 350 ^ long, 
9 {A thick at base, somewhat crooked, obtuse, black. Setse from 
near the base of the perithecium shorter, 135 [x. Perithecia glo- 
bose, smooth, 150 [x in diameter, originally flat and surrounded 
by radiating hyphge, at maturity globose. Asci evanescent. 
Spores 4-septate, 36 to 40 by 16 ix. 

Group number 3111.3221, 

On Sterculiacese : Pterospermum obliquum. Agricultural Col- 
lege, Laguna Province, Luzon, September 10, 1930, No. 498. 

Though similar to M. pterospermi, the only other species 
recorded as upon the Sterculiaceae, in formula this differs from 
it in its thin, not black and dense colony and in its regular 
head cells as well as in length of setse. 

MELIOLA PISONIICOLA sp. nor. Fiff. S, d. 

Colony 2 to 10 mm in diameter, thin, diffuse, indefinite. My- 
celium 6 to 7 (Jt. thick, somewhat crooked. Spot none. Capitate 
hyphopodia alternate. Stalk cell short, 3 {x, head cell subglobose 
to ovate. Mucronate hyphopodia ampulliform, short, thick. 

Perithecial setse none. Mycelial setse to 185 [a long, 6 to 7 
[x thick at base, simple, slightly curved, obtuse. Perithecia glo- 
bose, smooth, 70 to 80 ^ in diameter. Asci evanescent. Spores 
4-septate, 23 by 10 (jl. 

Group number 3111.2111. 

On Nyctaginaceae : Pisonia, San Jose to Balete Pass, Nueva 
Ecija Province, Luzon, January 10, 1931, No. 1813. 

This differs decidedly from M. pisoniae Stevens, which has op- 
posite capitate hyphopodia and a very dense and black colony. 

MELIOLA BRIDELIAE sp. nov. Fisr. a, e. 

Colony thin, circular, 3 to 10 mm in diameter, epiphyllous, in- 
conspicuous. Mycelium irregular, crooked, thin, 4 to 5 (x. Spot 
none. Capitate hyphopodia unilateral or alternate, not crowded. 
Stalk cell short, 2 to 3 [/., head cell cylindrical or oblong, standing 
out at a right angle or leaning abruptly forward. Mucronate 
hyphopodia ampulliform. 

Perithecial set^e none. Mycelial setse few, scattered, 580 to 
800 [X long, obtuse, simple, slightly curved, 6 to 7 [x thick at base, 
discal set« few, 4 to 8 at the base of each perithecium, 250 [k 
long. Perithecia globose, smooth, 116 to 120 \k in diameter. 
Asci evanescent. Spores 4-septate, 37 by 13 [x. 

70 Stevens and Roldan: Philippine Meliolinex 1935 

Group number 3111.3223. 

On Euphorbiacese : Bridelia stipularis. Naguilian Road, Ben- 
guet, Luzon, January 6, 1931, No. 1543. 

The colonies are very inconspicuous due to the scant mycelium 
loosely distributed over the leaf. Under the low power of the 
microscope a striking character is presented by the few long 
black mycelial setse. The two forms of mycelial setae are quite 
distinctive. This Meliola differs distinctly from all of the other 
species recorded upon the Euphorbiacese in its simple mycelial 
setse of two lengths, alternate, capitate hyphopodia, and setse 
more than 500 p. long. 


Group number 3111.2121. 

In specimens Nos. 801, 1714, 1723, 1061, on Sida acuta, each 
perithecial disk bore upward of twelve setse and no setse at all 
were seen on the mycelium itself. The same condition we find 
in Fungi Malayana No. 255. In case the mycelial setse are con- 
sistently absent, only the discal being present, the formula should 
be 3111.2121. If the mycelial setse are entirely absent and the 
setse around the perithecia are regarded as arising from the pe- 
rithecia rather than from the disks, then the formula would be 
3401.2121. and the fungus would belong to the genus Irenopsis. 
The fact that four varieties of Irenopsis have been described on 
the Malvacese is significant and, in view of the variability of 
Philippine collections of M. sidae, shows a close kinship between 
all of these forms. 


Group number 3111.4221. 

The Philippine material has a colony that is considerably more 
tenuous than the senior author's own collection from Panama 
and there are minor differences in capitate hyphopodia, but it 
seems well to regard the two as identical. 


Group number 3111.4221. 

On Tabernaemontana sp. San Jose-Balete Pass Road, and 
Muiioz, Nueva Ecija Province, Luzon. 

This appears to be the first record of this fungus in the Phil- 

MELIOLA MYCETIAE Stevens, «p. nor. Fiir. 3, f. 

Colonies amphigenous, small, 1 to 2 mm in diameter, black, of 
medium density. Mycelium dark, thick, 7 to 8 [x irregular to 

56, 1 Stevens and Roldan: Philippine Meliolinese 71 

crooked. Spot none. Capitate hyphopodia alternate, slanting 
forward. Stalk cell long, 6 to 8 [j., head cell ovate, regular. Mu- 
cronate hyphopodia ampullif orm. 

Perithecial setae none. Mycelial setse usually scant except 
near the perithecia, simple, obtuse, 190 to 205 \l long, very slight- 
ly undulate. Perithecia globose, smooth, 145 to 175 [j. in diam- 
eter. Asci evanescent. Spores 4-septate, 37 to 40 by 13 to 16 ^. 

Group number 3111.3221. 

On Rubiacese: Mycetia javanica. Kennon Road below Camp 
3, Benguet, Luzon, January 8, 1931, No. 1669. 

This appears to be the first record of Meliola upon Mycetia. 
This species is more closely related to M. eveae, M, ouroupariae, 
and M. psychotriae than to the other very numerous species of 
the Meliolinese that are recorded on the Rubiacese. Ours differs 
from M. eveae in the character of the mycelium, which in M. 
eveae is thinner and paler than in the present species. The 
capitate hyphopodia in M. eveae are also shorter. 

Meliola psychotriae has mycelium and hyphopodia of lighter 
color. Meliola ouroupariae differs very decidedly in its great 
abundance of setae, and in the small, nearly globose head cells. 
The present form differs also from each of the three mentioned 
above in the undulate character of the mycelial setse, 

A common character of many of the Meliolinese occurring upon 
the Rubiacese is that the capitate hyphopodia lean forward, to- 
ward the distal end of the mycelium that bears them, making an 
angle of approximately 90 degrees with the supporting hypha. 
This character is typically shown in the present species as well 
as in M. psychotriae. Another character common to the forms 
on the Rubiacese is that the head cells are entire, usually ovate. 

Group 10. Formula 3111. Set» obtuse or acute. 


Group number 3111.422|. 

On Burseracese: Canarium villosum. Mount Maquiling, La- 
guna Province, Luzon, July 22, 1930, No. 83. 

Comparison of the present collection with a specimen No. 
23877 of the Philippine Bureau of Science and 363 and with 
Fungi Malayana No. 547 convinces us that the determination 
of the above specimen is correct. We note, however, three 
important differences between this specimen and Sydow's de- 

1. Sydow describes the fungus as epiphyllous, while the pres- 
ent specimen shows it to be amphigenous. 

72 The Philippine Journal of Science isss 

2. Sydow makes no mention of deform mycelial setae, while 
we find, as well as the long mycelial setae that he describes, also 
other mycelial setae which occur only near the bases of the 
perithecia, these setae being few and short, 140 to 150 [x. 

3. The length of the mycelial setae is given in Sydow's de- 
scription as 300 to 550 |x, while in our specimens they reach 730 
(I. in length and are rarely below 500 [x. 


Group number 3111.3222. 

Two species of Meliola of nearly the same formula — ^namely, 
M. stenospora Winter 3111.4222 and M. piperina Sydow 3111. 
8222 — are recorded as on Piper in the Philippines. Meliola ste- 
nospora was described as with "hyphopodia promaxima parte 
late pyriformia, varie lobata." 

Sydow says of M. piperina, ''ausgezeichnet durch die stark ge- 
lappten Hyphopodien. Mit M. stenospora Wint. nachst ver- 

The senior author is unable to examine a type specimen of M. 
stenospora, but he has specimen No. 770 of the Bureau of Science, 
also a Kew specimen from Uganda, determined by Miss Wake- 
field as this species. Comparison of these with specimens of 
M. piperina Fungi Malayana No. 367 and Bureau of Science 
No. 23749 shows no differences. 


Group number 3111.4223. 

On OpUsmenus undulatifoUus. 

In some specimens the setae measure up to 540 [a thus making 
the last figure of the formula 3 instead of 2 as is indicated by 
the original description. The mycelium usually parallels the 
veins in the way common on species growing upon grasses. 


Colony 2 to 5 mm in diameter and circular or irregularly 
diffuse, spreading, thin, or coalescing and covering a large por- 
tion of the leaf. Mycelium 4 to 6 [jl thick, nearly straight. Spot 
none. Capitate hyphopodia alternate, leaning slightly forward, 
not crowded, about 30 to 35 [x apart. Stalk cell short, 3 to 4 [i, 
head cell ovate, regular. Mucronate hyphopodia ampulliform. 

Perithecial setae none. Mycelial setae 219 to 500 ]^ long, 7 (& 
thick at base, crooked, acute, simple. Perithecia globose, nearly 
smooth, 130 to 146 [a in diameter. Asci evanescent. Spores 4- 
septate, 30 to 33 by 13 to 14 ^. 

Group number 3111.3222. 

56,1 Stevens and Roldan: Philippine Meliolinese 73 

On Euphorbiacese: Ichnocarpus volubilis. Cuenca, Batangas 
Province, Luzon, September 28, 1930, No. 722a (type) ; Nagui- 
lian Eoad, Benguet, Luzon, January 5, 1931, No. 1465. 

This species differs from M. ramosii in having shorter and ob- 
tuse setse; from M. longispora in that the hyphopodia are ovate, 
not globose ; from M. gymnanthicola and its variety in colony ap- 
pearance and in setal tips ; from M. calUguajae in its small peri- 
thecia; from M. euphorbiae, M. sauropicola, and M. heveae in 
setal tips; from M. jatrophae in its longer setse; and from M. 
morbosa in not causing a spot. 


Colony thin, diffuse, epiphyllous. Mycelium 6 to 7 |x thick, 
dark, smooth. Spot none. Capitate hyphopodia unilateral or 
alternate. Stalk cell short, 3 to 4 jx^ head cell subglobose to 
pyriform or oblong, mostly oblong. Mucronate hyphopodia am- 

Perithecial setae none. Mycelial setse 525 to 730 ja long, 7 y, 
thick at base, black, crooked, acute at tip. Perithecia globose, 
190 pi, rough. Asci evanescent. Spores 3-celled, 40 to 44 by 
16 to 17 ix. 

Group number 3111.4223. 

On Myrtacese : Eugenia. Mount Maquiling, Laguna Province, 
Luzon, January 18, 1931, No. 1946. 

This species differs from the species of Meliola recorded as on 
members of the Myrtacese having a formula of 3111. as follows: 
From M. eugeniae, M. langiera, M. hawaiiensis, and M. laxa in 
length of setse and in its setse being acute. It differs from M. 
psidii and M. brasiliensis in many respects. 

An undetermined microthyriaceous fungus was very abundant 
on the lower sides of the leaves. 


Colony hypophyllous, black, velvety with setse, 5 to 10 mm 
in diameter. Mycelium somewhat crooked, dark, 6 to 7 (a thick. 
Spot none. Capitate hyphopodia alternate or unilateral, irregu- 
larly spaced. Stalk cell short, 3 to 4 jx, head cell ovoid to mostly 
cuneiform, regular. Mucronate hyphopodia ampuUiform, nar- 
row, short. 

Perithecial set« none. Mycelial setse most abundant near the 
perithecia, very black, straight, stiff, simple, acute, 10 \u thick, 
usually about 250 \j. long. Perithecia at first flattened, later glo- 
bose, black, smooth, 100 to 120 [a in diameter. Asci evanescent. 
Spores 4-septate, 47 to 50 by 16 |a. 

74 The Philippine Journal of Science isss 

Group number 3111.4221. 

On Euphorbiaceae : Glochidion sp. Naguilian Road, Benguet, 
Luzon, January 6, 1931, No. 1561. 

This differs in colony character and in hypophyllous habit 
from M. ramosii, from M, longispora in mycelial and setal char- 
acters, from M. gymnanthicola in its hypophyllous colonies and 
head cells, from M. jatrophae in hyphopodia, and from M. mar- 
bosa in spot character. 


Colony to 14 mm in diameter, irregular, black, amphigenous. 
Mycelium 6 to 7 \l thick, almost straight. Spot none. Capitate 
hyphopodia alternate, not crowded, but close, about 17 \k apart. 
Stalk cell short, 3 to 4 pi; head cell oblong, regular or slightly 
irregular. Mucronate hyphopodia ampuUif orm, numerous, most- 
ly opposite. 

Perithecial setae none. Mycelial setse very few, 200 to 220 pi 
long, 9 to 10 [1. thick at base, black, simple, acute. Perithecia 
globose, smooth. Asci evanescent. Spores 4-septate, 40 to 43 
by 16 to 17 [x. 

Group number 3111.4211. 

On Myrtacese: Eucalyptus. San Jose to Balete Pass, Nueva 
Ecija Province, Luzon, January 9, 1931, No. 1722. 

This differs markedly from all species of Meliola of formula 
8111. recorded on the Myrtaceae and from all species recorded on 


Group number 3111.3223. 

On Leguminosae: Spatholobus gyrocarpus. Mount Maquiling, 
Los Baiios, Laguna Province, Luzon, September 13, 1930. 

No Meliola has heretofore been recorded upon this host. 
Though the setae in our specimen in general are longer than in 
Sydow's specimen and as indicated in his description, it appears 
well to regard the two specimens as of the same species. 


Group number 3111.3222. 

Collections were made on Piper betle on which species of host 
it has not previously been reported. 


Colony 1 to 3 mm in diameter, black, moderately dense, epi- 
phyllous, mycelium 6 to 7 (x thick. Spot none. Capitate hypho- 
podia alternate or partly unilateral. Stalk cell short, 3 to 4 [x, 
head cell subglobose. Mucronate hyphopodia ampuUif orm. 

56, 1 Stevens and Roldan: PUlippine Meliolinese 75 

Perithecial setae none. Mycelial sete to 450 ^ long, 9 to 10 u 
thick at base, simple, acute, mostly near the center of the colony 
Penthecia globose, smooth, 120 ^ in diameter. Asci evanescent. 
Spores 4-septate, 40 by 17 pi. 

Group number 3111.3222. 

On Lecythidace^: Barringtonia sp. Quezon Forest Park, Ta- 
yabas Province, Luzon, November 30, 1930, No. 440. 

This is quite distinct from M, indica, the only species recorded 
on this family, in not having the capitate hyphopodia opposite. 


Group number 3111.4232. 

Four collections from Luzon agree with a specimen from the 
Bureau of Science and with Fungi Malayana No. 365, the distin- 
guishing character being the irregular shape of the head cells. 
The original description gives the head cells as globose or ovoid 
and does not mention the irregularity, which in the Philippine 
specimens is strongly pronounced. 



Group number 2111.4132. 

Our specimens collected in Laguna agree precisely with the 
type specimen, but since neither possesses perithecia, the deter- 
mination is questionable. 

Numerous of the senior author's Meliolineae collections are 
undeterminable due to such overgrowth by parasites as to pre- 
vent proper development of perithecia, setae, etc., or to scanty 
collections which did not give the requisite characters for de- 

It is obviously unwise to publish these with full descriptions 
and names, but since many of them represent new records on 
genera or families, it seems desirable to make record of their 
existence with a very brief description. In such cases the forms 
are designated by numbers, as follows: 

No 1. — Colony 1 to 5 mm in diameter, diffuse. Capitate hy- 
phopodia alternate, about 16 {/. apart, head cells oblong. 

On Ericacese: Vaccinium benguetense. 
. This is clearly not /. exilis or /. andromedae. 

No. 2. — Capitate hyphopodia alternate. Head cell subglobose 
to ovate, regular. 

On Thymelaeacese : Wikstroemia. Paete, Laguna Province, 

76 The Philippine Journal of Science i936 

Only two species of the Meliolinese are recorded on this fa- 
mily and neither of these from the Philippines. 

No. 8. — Capitate hyphopodia alternate, head cell oblong or ir- 
regular, mycelial setse 400 to 600 p. long, simple. 

On AnacardiacesB : Semecarpus sp. Mount Maquiling, La- 
guna Province, Luzon, No. 1921. 

This is a very characteristic form. It is not M. semecarpi, the 
only species recorded on this family from the Philippines, which 
has longer setse and many fewer capitate hyphopodia. 

No. i. — Capitate hyphopodia alternate, head cell irregular. 

On Pandanacese: Pandanus sp. Louisiana, Laguna Province, 
Luzon, September 21, 1930, No. 676. 

No species of the Meliolinese has hitherto been recorded on 
the Pandanacese in the Philippines, though one species, M. panr 
dani, is recorded from Borneo which is perhaps identical with 
this number. 

No. 5. — Capitate hyphopodia alternate, head cell subglobose 
to slightly irregular, very regularly spaced, colony very thin. 

On Sterculiacese : Pterospermum niveum. 

No species is recorded on this host from the Philippines, only 
four species are recorded for this family in the world. 

No. 6. — Capitate hyphopodia alternate, head cell ovate, regular. 

On Anonaceae: Papualthia lanceolata. Mount Maquiling, La- 
guna Province, Luzon. 

Though numerous species are known on the Anonacese, none 
has heretofore been noted upon this genus. It was collected 
on Mount Maquiling several times but in all cases very heavily 
overgrown by a Helminthosporium, so heavily that the presence 
or absence of setse could not be determined nor did perithecia 

No. 7. — Capitate hyphopodia numerous, alternate, head cell 
oblong or irregular, perithecia globose, smooth, colony dense. 

On Sapotacese: Palaquium sp. Mount Maquiling, Laguna 
Province, Luzon. 

No. 8. — Capitate hyphopodia few, distant, mycelial setse more 
than 1,100 {x long, perithecia smooth, globose, colony very thin. 

On Sapotacese: Palaquium sp. Mount Maquiling, Laguna 
Province, Luzon. 

Nos. 7 and 8 occurred upon the same leaves but were too scant 
to be determined. They certainly are two very different spe- 
cies and differ from M. palaquii, the only Meliola that has been 

5«, 1 Stevens and Roldan: Philippine Meliolinese 77 

recorded on Palaquium. None on the Sapotace^ has sete as 
long as those of No. 7. 

No. P.— Capitate hyphopodia alternate, head cell clavate, irreg- 
ular, colony thin. 

On Urticacese: Oreocnide sp. Acop's, Benguet Subprovince, 

No Meliola has been recorded on Oreocnide. 

No. iO.— Capitate hyphopodia alternate, head cell clavate, 
irregular, mycelium crooked, mycelial setae probably absent. 

On Vitace^ : Leea philippinensis. Mount Maquiling, Laguna 
Province, Luzon. 

No Meliola has been recorded on this host. 

No. ii.— Capitate hyphopodia alternate, head cell ovoid, reg- 
ular, mycelium very crooked. 

On Gesneriaceae : Isanthea discolor. Balete Pass, Nueva Ecija 
Province, Luzon. 

None has been recorded on this genus. 

No. 12. — Capitate hyphopodia alternate, head cell ovate, 'my- 
celial setae few, 80 to 90 pi. long, obtuse. 

On Compositae: Elephantopus. Balete Pass, Nueva Ecija 
Province, Luzon. 

None has been recorded on this genus from the Philippines. 

No. 13. — Capitate hyphopodia alternate, head cell cuneiform 
or angular, mycelium crooked. 

On Euphorbiaceae : Neotrewia cumingii. Mount Maquiling, 
Laguna Province, Luzon. 

No Meliola has been recorded upon this genus. 

No. 14'. — Capitate hyphopodia alternate, head cell ovate, reg- 
ular, mycelial setae long, to 500 [x, simple. 

On Tiliaceae: Grewia. 

No Meliola has been recorded upon this host genus hitherto. 

No. 15. — Capitate hyphopodia alternate, head cell ovate, reg- 

On Proteaceae : Helicia sp. 

No Meliola has ever been reported on Helicia. 

No. 16. — Capitate hyphopodia opposite, head cell conic-cylin- 
dric, mycelial setae simple, to 450 [x long. 

On Meliaceae: Dysoxylum. San Jose to Balete Pass, Nueva 
Ecija Province, Luzon. 

]S!o. 17. — Capitate hyphopodia alternate, head cell very large 
and irregularly lobed. 

78 The Philippine Journal of Science 

On Symplocacese : Symplocos. Mount Maquiling, Laguna 
Province, Luzon. 

No Meliola has been recorded for this family except the very 
questionable M. amphitricha, which in any event could not be 
the present species. 

No. 18. Fig. 3, i. — Colony circular, epiphyllous, 3 to 7 mm 
in diameter, usually heavily covered by an Arthrobotryum and 
sometimes by a Helminthosporium as well. Mycelium smooth, 
6 to 7 |i. thick, loosely distributed ; that is, not dense. Spot none. 
Capitate hyphopodia alternate. Stalk cell short, 2 to 3 [x, head 
cell rarely subglobose, usually irregularly angled. Mucronate 
hyphopodia ampulliform. 

Perithecial setae, mycelial setae, perithecia, and spores not 

On Chloranthacese : Chloranthus officinalis. Mount Maqui- 
ling, Laguna Province, Luzon. 

No species of Meliola has heretofore been recorded on any 
member of the Chloranthaceae ; though the present species was 
collected many times on Mount Maquiling at different elevations, 
in no instance was even a single colony seen that was not heavily 
overgrown by an Arthrohotryum, so heavily indeed that the Me- 
liola mycelium was almost completely obscured and setae and 
perithecia were entirely suppressed. A rather extraordinary 
feature of these colonies is that occasionally in the midst of my- 
celial strands that are heavily parasitized are a few that are 
not parasitized at all. The dense sheathing of the Meliola by 
the Arthrobotryum mycelium is very characteristic and it is only 
in rare bits that the Meliola mycelium and its hyphopodia can 
be seen. 

In the absence of the usual Meliola characters, the chief dis- 
tinctive character of this species is that of its being so heavily 
parasitized, and its host relation. The fact that this Meliola is 
so heavily parasitized, while most other species of Meliola in 
the vicinity are not, is strong evidence of the biologic spe- 
cialization of these fungi growing parasitic on Meliola. 



Fig. 1. species of Meliolaf Irenopsis, and Irenina, 

a, Meliola saurauiae sp. nov., mycelium habit, low power, myce- 
lium showing detail, and spore, high power. 

6. Irenopsis benguetensis sp. nov., mycelium showing capitate 
and mucronate hyphopodia, high power. 

e, Irenina acalyphae sp. nov., mycelium showing capitate and 
mucronate hyphopodia, high power. 

d, Irenina rubi sp. nov., mycelium showing habit, low power, 

and mycelium showing capitate and mucronate hyphopodia, 
high power. 

e, Irenina rubi Stevens var. angulata var. nov., mycelium show- 

ing capitate and mucronate hyphopodia, high power. 

/. Irenina angustispora sp. nov., mycelium showing habit, low 
power, and mycelium showing capitate hyphopodia, high 

flr. Irenina callicarpae sp. nov., mycelium showing habit, low 
power, and mycelium showing capitate hyphopodia, high 

h. Irenina thunbergiae sp. nov., mycelium showing capitate and 
mucronate hyphopodia, high power. 

i, Irenina sinuesa sp. nov., mycelium showing capitate and mu- 
cronate hyphopodia, high power. 
2, Species of Meliola. 

a. Meliola palaquii sp. nov., mycelium showing capitate and mu- 
cronate hyphopodia, high power. 

6. Meliola micromeli sp. nov., mycelium showing capitate and 
mucronate hyphopodia, high power, and setas showing apices, 
high power. 

c. Meliola benguetensis sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power, and setse showing teeth, 
high power. 

d. Meliola pistaciae sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power, and set« showing teeth, 
high power. 

e. Meliola themedae sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power, and seta showing tip, 
high power. 
/. Meliola nepheliicola sp. nov., mycelium showing capitate and 
mucronate hyphopodia, high power, and setas showing 
apices, high power. 


80 The Philippine Journal of Science 

g, Meliola symphoremae sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power. 
h. Meliola pisoniae sp. nov., mycelium showing capitate and mu- 

cronate hyphopodia, high power. 
t. Meliola rubi sp. nov., mycelium showing capitate and mucro- 

nate hyphopodia, high power. 
8. Species of Irenopsis and Meliola, 

a. Meliola illigerae sp. nov., mycelium showing capitate and mu- 

cronate hyphopodia, high power. 
6. Meliola meliacearum sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power. 

c. Meliola pterospermicola sp. nov., mycelium habit, low power, 

and mycelium showing capitate and mucronate hyphopodia, 
high power. 

d. Meliola pisoniicola sp. nov., mycelium showing capitate and 

mucronate hyphopodia, high power. 

e. Meliola brideliae sp. nov., mycelium showing capitate hypho- 

podia, high power. 

/. Meliola mycetiss sp. nov., mycelium showing habit, low power, 
and setae showing undulations, high power. 

g, Meliola agelaeae sp. nov., mycelium showing capitate and mu- 
cronate hyphopodia, high power. 

h, Meliola harringtoniicola sp. nov., mycelium showing capitate 
and mucronate hyphopodia, high power. 

L Meliola 18, mycelium showing capitate hyphopodia, high 

i, Irenopsis coronata var. philippinensis var. nov., mycelium 
showing habit, low power, and mycelium showing capitate and 
mucronate hyphopodia, high power. 

k. Meliola ganophylli sp. nov., mycelium showing capitate and 
mucronate hyphopodia, high power, and setas showing apices, 
high power. 

I, Meliola andropogonis sp. nov., mycelium showing capitate hy- 
phopodia, high power. 


By RiCARDO E. Galang 
Of the National Mvseum Division, Bureau of Science, Manila 


Many ethnographic studies on the different Philippine groups 
have been undertaken by eminent anthropologists and scientists, 
yet the Yogads have never been given proper attention, perhaps 
because of their small number and the similarity of their charac- 
teristics to those of the Gaddangs and the Ibanags. The litera- 
ture references available on the group are very limited, being 
confined to their dialect, geographic location, and their relation 
to other ethnic groups. Blumentritt in 1890 only mentioned 
their geographic location and their relation to the Christianized 
Gaddangs. Pardo de Tavera in 1901 analyzed the etymology of 
their tribal name. Worcester in 1906 mentioned them as a part 
of the Christianized Kalingas of western Isabela. Malumbres 
in 1918 likewise mentions their geographic location and the 
relation of their dialect to the Gaddangs. The present writer 
attempts to present in this short paper a study of their economic 
and social life, and their dialect. In 1931, when he was officially 
sent to Isabela and Nueva Vizcaya Provinces to purchase mu- 
seum specimens, he stayed with this group long enough to become 
acquainted with it ; and, having been given the necessary facilities 
for ethnographic study, he was able to familiarize himself with 
the economic and social life of the Yogads, and the peculiarities 
of their dialect. These studies form the subject of the present 


The term Yogad, or logad, is derived from ugad, an Ibanag 
word meaning elevated garden patches or seed plots. 

Se conoce con el nombre de yogades (yogad 6 gaddanes a unos monta- 
fieces que en la actualidad habitan en las vertientes de la Cordillera central, 
Hmite occidental de la provincia de la Isabela, desde la orilla izquierda del 
Rio Magat, termino del Rio Mercedes, hasta la jurisdicci6n de los pueblos 
cristianos de Itawes. 

289095 — « 81 

82 The Philippine Journal of Science 1935 

La etimologia de ambos nombres nos la da la lengua ibanag que tanto 
se parece al gaddan, resultando que gaddan y yogad tienen la misma sig- 
nificacion y origen, solo que el primero es nombre de lugar y el segundo 
de tribu, Ambos derivan del radical gad cuya forma mas antigua ugad, 
existe tambi^n en ibanag y significa "sementera alta". Esta raiz, seguida 
del sufijo an, quiere decir "lugar de la sementera alta" como pay aw vimos 
que queria decir "sementera baja o de regadio". Los habitantes del gaddan, 
llamados hoy asi con este nombre propio de personas, que es yogad for- 
mado con el prefijo i cuyo significado conocemos y el radical ugad igual 
a gad que hemos examinado.^ 

It is said that the Yogads, who regard themselves as offshoots 
of the Ibanags, originally came from the town of Enrile, formerly- 
called Cabug, which was established in 1742 in Cagayan Prov- 
ince. Due to the crowded conditions of their native town and 
the difficulties encountered in the struggle for existence, this 
sturdy, peace-loving, and adventurous people decided to migrate 
to Isabela Province, to find a new home and to seek their for- 

Worcester ^ states that the Yogads are a part of the Chris- 
tianized Kalingas of western Isabela. These Kalingas are called 
Gaddangs, and were Christianized by missionaries. 


The Yogads were at one time confined to the level land of the 
town of Difun, which was established in 1706. The principal 
towns which they now occupy are Camarag, the oldest town in 
Isabela, established in 1753, and Echague, Angadanan, Santiago, 
and Jones, formerly Cabanuangan. 


The present number of the group speaking the Yogad dialect 
is conservatively estimated at 8,000. This number represents 
those who are regarded as being of pure stock. A few of them 
have migrated to neighboring provinces where they intermarried 
with the other Christian groups. 


The Yogads are predominantly of the Indonesian type with a 
slight admixture of Negrito and Chinese blood. They vary from 
reddish brown to dark brown in complexion. They are round- 
headed, with straight, black hair, dark brown eyes, and a nose 

^Pardo de Tavera, T. H., Etimologia de los Nombres de Razas de 
Pilipinas. Manila (1901) 14. 

• Philip. Journ. Sci. 1 (1906) 818. 

56,1 Galang: The Yogads of Isabela 83 

of medium breadth with a low bridge. They are of ordinary 
size and have unusually regular features. Like other Philippine 
peoples, they are practically beardless. 


The Yogad culture, like that of the other Christian groups 
of the lowlands, is tinctured by Spanish civilization, although 
the people retain many of their primitive customs and beliefs. 


Agriculture. — Formerly the Yogads practiced the kaingin sys- 
tem of agriculture, planting in cleared forest areas where the 
soil is loose and fertile. Under this system the small trees and 
shrubs are cut away and burned, and the large trees are killed. 

According to the census of 1918 the climate of the plains occu- 
pied by the Yogads is very favorable to the growth of tobacco. 
The northeast monsoons bring heavy rains, which wash down 
the fertile mountain soil into the rivers that deposit the silt on 
the plains. In this manner the tobacco fields are fertilized every 

The principal product is tobacco, which is grown in large 
quantities. The variety of tobacco raised in this region is con- 
sidered one of the best in the Philippines, and, as one of the 
principal articles of export, constitutes the wealth of the people. 
The corn crop is the object of considerable care on the part of 
the natives, as it constitutes their principal food supply when the 
price of rice is high. Rice, sugar cane, coconuts, and coffee, 
grow almost without the care of the planter, A few cattle are 
raised. The forests are rich in valuable timber, such as molave, 
ipil, narra, and camagon. 

Hunting and fishing. — Hunting is carried on to a considerable 
extent. Wild pigs, carabaos, and deer are caught with snares, 
traps, and barbed spears that are similar to those of the Bontoc 
Igorots. The Sierra Madre Mountains are the hunting places of 
the Yogads. 

Cagayan River has an abundant supply of fish, and fishing is 
one of the principal occupations of the people. Fish are caught 
with nets, lines, and traps, as well as with small barbed spears, 
and bows and arrows. Murrel fish constitute the principal 

Basketry.— The Yogads do considerable basket work. The 
principal materials used are bamboo, rattan, and nito. Four 

"Census of the Philippine Islands: 1918 1 (1920) 167. 

84 The Philippine Journal of Science ma 

types of weave are employed; namely, "in and out" or cloth 
weave, twilling or sawale weave, hexagonal weave, and coiling. 
Some of their baskets are patterned after those of the Ilocanos 
and pagan tribes of northern Luzon. 

Weaving of textiles and the making of pottery are unknown 
among the Yogads, who purchase most of their clothing from 
Ilocano and Chinese traders, and their pots and stoves from the 
Tinggians and Ilocanos. 

Transportation. — The Yogads have transportation by land and 
water. Their typical land transportation is the cart drawn by 
a carabao or cow. Riding on horseback is also common on 
mountain trails. 

Their typical water transportation is the raft, or gakit, and 
the dugout canoe, or abdng. The gakit, which is about 6 meters 
long, is made of mountain cane, or buhu, lashed together by 
rattan. It is used for transporting across Cagayan River mature 
tobacco leaves placed in big baskets called tangkal. The dugout 
canoe is used for fishing and passenger transportation. 

Dwelling. — The Yogad dwelling is a structure of wood, bam- 
boo, or mountain cane, with a thatched or bamboo roof. This 
kind of roofing, or camd, must be in imitation of the Ilocanos 
who commonly use it in roofing their houses. The sides or walls 
are made of sawale or mountain cane split into halves, which are 
placed so as to overlap one another. The floor is made of wood, 
sawale, or bamboo. The houses have only one partition and are 
raised 3 to 4 feet above the ground. 

Dress and ornaments. — ^Yogad dress is similar to that of other 
Christian groups. The men wear shirt and trousers and the 
women camisa and skirt. 

Yogads are not much given to personal ornamentation. Some 
of the men have tattoo marks on the arms and hands. Necklaces 
of beads and coconut shells, earrings, and finger rings are also 
worn to some extent. 


Music. — The Yogads, like the other Christian Filipinos, are 
very fond of music. Social gatherings and entertainments are 
always marked by singing and dancing. 

The typical musical instrument of the Yogads is the cinco- 
cinco, a small five-stringed wooden guitar, which is used to ac- 
company songs and dances. Other musical instruments are the 
tal'lelet and the caralat, or bamboo rattles. These are percus- 
sion bamboo musical instruments, which are played at night 

s«'i Galang: The Yogads of Isabela 35 

during Holy Week. A band contest for these musical instru- 
ments is held by the different barrios at a designated place. 

Dances.— The Yogads have four typical dances; namely, mas- 
cota, a la jota, laurente, and balamban. The first three dances 
are similar to each other although the music differs. They are 
danced, to the accompaniment of a cinco-cinco or an accordion, 
by a man and a woman facing each other. The rhythmic move- 
ment of the hands is emphasized. The balamban is also danced 
to particular music by a man and a woman, with special em- 
phasis on the rhythmic movement of the feet. The name of this 
dance originated from the name of a long fish, called balamban, 
that jumps in the water. 

Circumcision.— The Yogads, like some of the other Christian 
groups, practice circumcision. The boys are usually circum- 
cised at twelve to fourteen years of age. The foreskin or pre- 
puce of the penis is split with a sharp knife. Young guava 
leaves are masticated and the sap is applied as medicine. It 
takes ten to fifteen days to heal the wound. This practice is 
called banguit. 

Marriage customs. — Generally speaking, the Yogads prefer to 
marry among themselves. They seldom intermarry with other 
Christian groups. 

Formerly it was customary for the parents to make arrange- 
ments for the marriage of their children. Now it is left to the 
discretion of the boy and the girl. As soon as the boy wants 
to marry, his parents together with a spokesman go to the girl's 
parents and ask for her hand. They bring with them a drink, 
usually gin. 

On the third day the spokesman of the boy's parents is told to 
come again to the house of the girl for the decision of the girFs 
parents. If the decision is favorable, the parents and relatives 
of both parties meet on the third day to make the final arrange- 
ment for the wedding. 

Then the landai takes place. This is a feast prepared by both 
parties prior to the real wedding feast. A sort of poetical joust 
between the representatives of the boy and the girl usually 
takes place. While the contest is being held, the girl prefers to 
stay in the kitchen. The representative of the girl usually yields 
to the representative of the boy and then they all partake of the 

A band of music usually follows the bride and the groom on 
their way to the church where the wedding ceremony is to be 

86 The Philippine Journal of Science 1935 

After the wedding, the friends and relatives of the couple are 
entertained with a dance. Then follows what is called gala, or 
the giving of money to the couple. Two plates, one for the bride 
and the other for the groom, are placed on the table. The 
relatives of the bride and the groom put their money on the 
respective plates of the couple. 

After the gala, the so-called dal-lut takes place. Only the 
relatives of both parties take part. It is usually performed at 
3 or 4 o'clock in the morning. Some of the male relatives join 
their hands and some of the female relatives do the same. The 
male relatives of the couple are allowed by common consent to 
kiss the female relatives without holding them. After the dal- 
lut the girl goes to the parents of the boy, and the boy to the 
parents of the girl. Then they live together as husband and 

Superstitious beliefs. — ^The Yogads are a very superstitious 
people, especially those who live in Barrio Capitan, Echague. 
They believe in evil spirits, and the Anitos, who are said to cause 
sickness, which, however, may be cured by an anting-unting, or 

The carangat, or evil spirits, dwell in certain places, such as 
bamboo, balete, and andarayan trees. When a person gets sick, 
an old man called maguimun is summoned to find out whether 
the sickness is caused by the evil spirits. He asks the sick per- 
son where he went or played to ascertain the cause of his sickness. 
If the sickness is caused by the evil spirits, he takes little pieces 
of buyo, rolled tobacco leaves, rice (malagquit) placed in a coco- 
nut shell, and a white chicken and goes to the dwelling place of 
the spirits. He communes with the spirits, shouting at them 
and telling them toieave the sick person and to return to their 
dwelling places for he has prepared food for them. If the sick- 
ness of the person is very severe, the maguimun cuts the left 
leg of the chicken and the right leg is tied with a cotton thread. 
Then he anoints the forehead of the sick person with the chicken's 
blood. It is believed that the sickness will be cured in this 

The mag-anito is another common superstitious belief among 
the Yogads. They go to the forest for three days' festival 
during the dry season, taking with them pigs and rice. The 
maponags, or women mediums, are the ones who commune with 
the anitos. Part of the pigs is offered to the anitos, and the rest 
is cooked for the members of the party. The share offered to 
the anitos is usually given to the maponag. This yearly three- 

66»i Galang: The Yogads of Isabela 87 

day festival has to be repeated seven times. It is believed that 
after seven years the members of the party will be immune from 
sickness caused by the anitos. During these seven years, the 
members are not allowed to eat sugar and garlic. Only salt is 
allowed to be used in their food ; nor are they allowed to repair 
their houses or use a white blanket or curtain. 

Burial ceremony. — Before burial, a white handkerchief is 
spread on the face of the dead person. Then those present pray 
and leave the dead, except for the one who is to remain to hold 
the four corners of the handkerchief. As soon as the latter 
believes that the spirit of the dead is already in the handkerchief, 
he takes it home and walks around the house. It is believed 
that on the third day the spirit of the dead will visit them and 
stay only under the house if such a ceremony is not performed. 

The members of the family are not allowed to stoop or look 
down when the coffin is being lowered into the grave, for it is 
the belief of the Yogads that he who looks down will also die. 

After burying the dead, the members of the family wash their 
hands and faces with water mixed with burned straw. 

For nine nights they pray for the spirit of their dead, every 
night offering food to the spirit. 


The Yogad dialect resembles quite closely Ibanag and Gad- 
dang. Malumbres ^ states that the Yogad dialect is but modified 
Gaddang. The Yogads have a peculiar intonation in speaking. 
They say their prayer in Ibanag. They have no printed litera- 
ture or periodicals. 

It has been observed that the majority of the Philippine 
dialects lack labial fricatives.^ However, this is not true in 
the case in the Yogad dialect. Like Ibanag and Gaddang, it 
possesses the surd aspirant /, a sound which is very common 
in this dialect. The original p becomes / when it is immediately 
followed by u. For example, fire, afuy; leg, uffu; heart, futu; 
ten, tafulu. Judging from the material at hand, the sonant 
aspirant v is not used in the Yogad dialect. The original 6 
is persistent. For example; moon, bulan; pig, buhay; thousand, 

*Historia de Cagaydn. Manila (1918) 14. 

* Conant, C. E., "F" and "V" in Philippine languages, Philippine Ethnol- 
ogical Publications 5 (1908) 135. 


The Philippine Journal of Science 



I have selected some of the commonest terms from the work 
of Otto Scheerer ^ to show the affinity of the Yogad dialect to 
its sister groups of dialect. 

Common terms showing the c 

iffinity of Yogad to Ihai 

lag and Gc 











\mata tal Ungui-t 
























































































H puto 



































• Scheerer, Otto, the Batan dialect as a member of the Philippine group 
of languages, Philippine Ethnological Publications 5 (1908) 22. 


Galang: The Yogads of Isabela 


Common terms 

showing the offinity of Yogal to Ibanag and Gaddan^— Ctd. 


























































































tafulu tatta 


tafulu tdtta 



dua fulu 


one hundred 




From the common terms mentioned above, it is evident that 
the Yogad dialect closely resembles Ibanag and Gaddang. In 
Yogad and Ibanag the accent appears frequently on the second 
syllable of the word. For example : 

























Words not marked by an accent, are pronounced as in Yogad 
and Ibanag. For example: 














The Philippine Journal of Science 

In the Gaddang dialect the accent frequently appears on the 
first syllable of the word. For example : 































one hundred 


It is evident that the Yogad and Ibanag dialects have the 
same intonation in speaking, while that of the Gaddang is 


The Yogads belong to the group of Ibanag emigrants who 
settled in Isabela Province to improve their economic life. Their 
physical characteristics resemble those of the Christian Gad- 
dangs living in Magat River Valley. Their social life is quite 
distinct, for they retain many of their primitive customs and 
beliefs. Their dialect resembles quite closely the dialect groups 
spoken in Cagayan Province, the home of the Ibanag people. 


Plate 1 

Fig. 1. A typical Yogad house (binalai). Its walls are made of split 
mountain bamboo strips placed so as to overlap one another. 
2. The sledge basket (tangkal) used for hauling mature tobacco 

Plate 2 

Fig. 1. Typical fish spears of the Yogads. 

2. The circumcision instrument (pag-hanguit) , 

Plate 3 

Fig. 1. The bamboo rattle {tal4elet). 

2. The five-stringed wooden guitar (cineo-cinco) . 


il.xtAW*: TiiK Y«m:/\iis nr i^A.w.t,,\,'\ 

I I'll!!, IP. .Itjf i;n. s<-l. ;">«, N,,. I. 

./ ^■ 


<^\l.,\Nii; Thk Y«r,.\li>t i»f- lsABK!..\.1 ,hi\'K'-i Sii , .'*»;, \%i, I, 



V*Ah\SV.: TllR YiHlM*K ,>t.> lSAMKL\,| 

f Fmi.iF ,lf»! IL\. Sr 






By Canuto G. Manuel 
Ornithologist, Fish and Game Administration, Bureau of Science, Manila 


Two birds in the ornithological collection of the Bureau of 
Science attracted my attention. One proved to be a new sub- 
species of Cyomis. The other is a female of Prionochilus an- 
thonyi McGregor; the type of this dimorphic species is a male.^ 


Type — Female, No. 13347 Bureau of Science collection; Mount 
Crista at 800 meters, Cagayan Province, Luzon, Philippine Is- 
lands, April 6, 1929; collected by Francisco Rivera. 

Subspecific characters. — Upper surface resembling that of C. 
lemprieri (Sharpe), of Palawan, Balabac, and the Calamianes, 
more than that of any other species. However, the brownish 
olivaceous ^ tinge is more uniform throughout the upper surface 
than in C. lemprieri. Color of underparts distinct from that of 
C. lemprieri in not having white on chin and possessing lighter 
orange-buff on throat and upper breast. This subspecies also 
differs from C. lemprieri in the length of wings and tail. The 
remaining four complete feathers of the mutilated tail are 
shorter than any of the tail feathers of C. lemprieri from Puerto 
Princesa and Tanabog, both in Palawan, that were examined. 

This subspecies differs from C. beccariana simplex Blyth of 
Luzon and Marinduque in having no blue on the upper parts. 
Robinson and Kinnear ^ however, are of the opinion that the 
blue color of the upper surface shows a "considerable general 
variation." Tawny-olive band on forehead absent in C. b. sim- 
plex. Undersurface distinctly lighter. 

In accordance with the characters laid down by Chaseen and 
Kloss 4 for the banyumas group, ''crown and back brown, or 

^Philip. Journ. Sci. § D 9 (1914) 531. 

« Colors are from Robert Ridgway, Color Standards and Color Nomen- 
clature. Washington (1912). 
"Nov. Zool. 34 (1928) 231-261. 
*Bun. Raffles Mus. No. 2 (1929) 23-42. 


94 The Philippine Journal of Science 

grey-brown, in females; tail either brown or particolored blue 
and brown'* this bird is considered a new form of this group 
with its range in the highlands of northern Luzon. 

Description, — ^A band of tawny-olive on forehead, which grad- 
ually disappears over each eye; crown and hind neck deep 
grayish olive ; back and greater coverts dark citrine, basal half 
of these feathers grayish brown ; rump ^ dark citrine ; upper 
tail coverts Sudan brown; tail Sanford's brown; primaries and 
secondaries blackish brown, their outer webs with snuff brown 
edges; tertiaries olive-brown. Chin, throat, and fore breast 
orange-buff, becoming brownish buff on breast; abdomen with 
a circular patch of grayish white, about 20 millimeters in 
diameter; flanks dark citrine. Under tail coverts (mutilated) 
grayish white. A pair of loral bristles white. 

Wing, 78 millimeters ; culmen from nostril, 11 ; tarsus, 18. 

This subspecies is named for Mr. Richard C. McGregor, ornith- 
ologist of the Philippine Government, who, for the last thirty- 
three years, has contributed much to the knowledge of the Phil- 
ippine avifauna. 


Type of female. — No. 28514 Bureau of Science collection; 
Mount Tabuan, at about 1,500 meters, Cagayan Province, Luzon, 
Philippine Islands, May 18, 1929 ; collected by Francisco Rivera, 

Entire upper surface and sides of head olive-green, which is 
more intense on back, rump, and upper tail coverts. Primaries 
and secondaries dark brown with little olivaceous gloss. Outer 
webs of quills with olive-green margins, those of the inner webs 
light brown. Primary coverts dark brown, other coverts olive- 
green. Chin sparsely covered with antrorse white feathers; 
throat gray; sides of breast yellowish citrine; sides of abdomen 
and thigh olive lake; a line of primuline yellow along middle 
breast and abdomen, including entire under tail coverts. 

Bill black above and matt ivory yellow below; legs, feet, and 
claws Brussels brown ; soles pale yellow, same as in male. 

Length, 87 millimeters ; wing, 59 ; culmen, 10 ; tail, 29 ; tarsus, 

The mate (Bureau of Science collection, No. 28513) of this 
female was also obtained. 

'Parts are mutilated. 


Plate 1. Cyomis banyumas mcgregori subsp. nov., female. 


\1 \ M n . .\i u itiia*-^ I t;u\i \%>Ki»!fi, . l^/t.^. 

! |-|r;U}' .|o| j;\ S( !.. ,">!;, Xi 


The Philippine 
JouKNAL OF Science 

Vol. 56 FEBRUARY, 1935 No. 2 



Of the Philippine Department of Agriculture and Commerce, Manila 

OPHIOGLOSSUM RAMOSII Copeland sp. nov. Plate 1. fi^s. 1 and 2. 

0. simplici simile, majus; caule parva, ut videtur subterranea; 
frondibus solitariis vel 2-3 fasciculatis, usque ad 45 cm altis et 
4 cm latis sine discrimine stipitis et laminae, vestigio segmenti 
sterilis nullo vel dubio ; spica apicale, 5 cm longa, 5 mm lata, lan- 
ceolata, sporangiis lateralibus, baud interruptis, parvis. 

Camiguin (Mindanao), Mount Mahinag, in mossy forest; Bu, 
Set. H771 Ramos, April, 1912. 

This remarkable plant has been held unnamed long enough, 
in the hope of more ample collection. It is now a pleasure to 
give it the name of its discoverer, Maximo Ramos, who lost his 
life in the field. May 8, 1932, after nearly thirty years of zealous 
and productive botanical collecting for the Philippine Govern- 

CYATHEA BIPINNATIFIDA Copeland sp. nor. Plate 2. 

Trunco ignoto ; stipite 15 cm longo, gracile, deorsum paleis 5- 
12 mm longis basi ultra 1 mm latis apice setaceis fulvo-griseis 
iridescentibus vestito, sursum pilis fulvis 3 mm longis dense 
vestito ; f ronde 25 cm longa, 10 cm lata, acuminata, basi truncata, 
bipinnatifida, pinnis infimis subdeflexis, rhachi setosa setis basi 
incrassatis; pinnis sessilibus, lineari-ellipticis, majoribus 5 cm 

* The seventh paper with this title appeared in Philip. Journ. Sci. 40 
(1929) 291-315. 

289853 97 

98 The Philippine Journal of Science 1935 

longis 1 cm latis, obtusis vel acutis, basi truncatis, pinnatifldis, 
costis setis deorsum dense et squamulis paucis subbuUatis api- 
culatis intersparsis vestitis; segmento infimo pinnarum inferio- 
rum fere libero late elliptico, aliis late oblongis, 3 mm latis, sub- 
f alcatis, apice late rotundatis, integris, herbaceis, glabris ; venis 
ca. 4-paribus, simplicibus; soris medialibus, indusiis ut videtur 

Basilan, east of Cumalarang River, Bu. Sci. 16208 Reillo, 
September, 1912. 

This fern has spent the past two decades among the undeter- 
mined specimens of Dryopteris. The palese are those of Cych 
thea, and the sporangia fix its position there; but I know no 
related species. 

CTATHEA UMBROSA Copeland sp. noT. Plate 3. 

Stipitis basi paleis nitidis aut albidis aut laete f ulvis usque ad 
5 cm longis et 4 mm latis apice in setam longam setuliferam aut 
albam aut brennescentem protensis; pinna longe (8 cm) stipi- 
tulata, 75 cm longa, fere 30 cm lata, rhachi baud dense tuber- 
culata, fere glabra; pinnuHs permultis, infimis solummodo 
pedicellatis, usque ad 15 cm longis, 22 mm latis, serrato-acumi- 
natis nisi apices versus pinnatis, costis inferne squamulis albis 
integris plerisque ovatis 1 mm longis sparsis ornatis; pinnulis^^ 
integris, acutis, plerisque curvis, papyraceis, inferne glaucis, cos- 
tulis minute et sparse squamuliferis; venis inferne conspicuis, 
plerisque bis furcatis; soris inframedialibus, nudis, apices versus 

Luzon, Sorsogon Province, Mount Bulusan, Elmer 16588, July, 

This was distributed with my identification as C. lepifera, 
but new study satisfies me that it is too distinct to justify this. 
It is distinct in ampleness of frond, in thinness, in glaucousness, 
and particularly in nakedness. 

Elmer 17070, from the same place, conforms satisfactorily to 
our cotype of C. lepifera, from the same general area. It is like 
it in denseness of tubercles and palese, and in firmer texture, 
but is more ample. The palese of the base of the stipe are golden 
in mass color, each being bordered by a narrow dark-golden line. 

CYATHEA PTERIDIOIDES Copeland sp. nov, Plate 4. 

Trunco 3.5 m alto, stipite ignoto, fronde 1 m (vel ultra) longa, 
70 cm lata, rhachi dorsum 1 cm crassa sub paleis imbricatis late 
lanceolatis 1 cm longis albidis vel medio laete rufis dense fur- 

66,2 Copeland: Philippine Ferns, VIII 99 

furacea, sursum inferne tuberculis globosis nitentibus 0.5 mm 
diametro horrida, apicem versus rhachibusque pinnarum squamis 
tenuissimis ovatis laceris albidis profunda immersis ; pinnis sti- 
pitulatis medialibus 35 cm longis, abrupte brevi-acuminatis, pin- 
nulis contiguis, subsessilibus, 10 cm longis, basi 18 mm latis, 
acuminatis, deorsum pinnatis, costis deorsum inferne dense pa- 
leaceis, sursum nudis, pinnulis" resp. segmentis 3 mm latis, 
obtusis marginibus integris revolutis, costulis deorsum paleatis, 
venis inferne praestantibus, furcatis, lamina nuda, rigide charta- 
cea; soris inframedialibus, paleis protectis, indusiis nullis. 

PANAY, Antique Province, Culasi, altitude 1,500 meters, Mc- 
Gregor 6061, June, 1918. 

The color, texture, and dissection give the herbarium spec- 
imen the appearance of Pteridium. Another representative of 
the small group of species known by the Javan C. tomentosa, C. 
crinita of Ceylon, and C. lepifera of 'Luzon (Camarines Sur) : 
like the last in texture, but %vith broader pinnules and segments, 
less completely tripinnate, much scalier on the major axes, and 
naked toward all apices. 

DRTOPTERIS SQUAMIPES Capeland sp. nov. Plate 5. 

Dryopteris D. viscosae affinis pinnis deorsum sensim decres- 
centibus; caudice erecta, valida, breve; stipitibus dense fascicu- 
latis, 5-7 cm altis, pervalidibus, atropurpureis, paleis ovatis acu- 
minatis 5 mm longis fuscis dense vestitis; fronde 25-35 cm alta, 
media altitudine 6 cm lata, acuminata, deorsum longe attenuata, 
vix bipinnata, rhachi ubique dense ferrugineo-pilosa deorsum 
quoque inferne squamosa ; pinnis medialibus 3 cm longis, subses- 
silibus, basi 1 cm latis, acutis, vix ad costam pinnatifidis, costa 
supeme brevisetosa, inferne et pilis et paleis minutis angustis 
obsita; segmentis ellipticis, apice rotundatis, crenulatis, utraque 
facie sparse piluliferis, atroviridibus, in herbario papyraceis; 
venis ca. 5-paribus; soris magnis, 3- vel 4-paribus, faciem infe- 
riorem fere obtegentibus, indusio persistente, f ulvo, suborbiculare 
vel elongato, bullato, nudo. 

Mindanao, Bukidnon Province, Mount Lipa, Bu. Set. S8525 
Ramos and Edano, "on dry, mossy forest near summit, altitude 
6,600 feet." 

In form of frond like many species of cooler lands, relatives 
of Thelypteris; but more nearly related, I believe, to D. viscosa. 
Growth on an exposed mountain top might cause that species to 
produce shorter, stouter, and more scaly stipes, but could hardly 

100 The Philippine Journal of Science 1935 

be responsible for the gradual dwarfing of the lower pinnse. 
Four or more pairs of these are deflexed, and the lowest may be 
15 mm long. 


Dryopteris Preslii CHRIST, in Philip. Journ. Sci. § C 2 (1907) 214, 
non (Baker) O. K. 

I have already [Philip. Journ. Sci. 40 (1929) 294] called at- 
tention to Christ's misinterpretation of this species, and am now 
able to place the plant he had in hand, Cuming S5i from Bohol 
as represented in the Philippine National Herbarium. This is 
a dwarf specimen of Dryopteris dissecta (Forster) 0. K., the 
largest fronds 6 by 4 cm, undissected in correlation with its size, 
but fertile. There is no uncertainty as to the affinity, and I am 
not now disposed to construe it as genetically distinct. Another 
dwarf, less certainly representing D. dissecta, is Bur. Sci. JfllSi 
Ramos, Coron Island, Palawan. The specimens of Cuming S54 
in other herbaria, called Lastraea spectabilis Presl, Epim. 38, 
D. syrmatica, etc., must be very unlike ours. 

It may be observed that, if the species be construed closely, 
several are included under D. dissecta. The typical Polynesian 
plant has pubescent veins and almost naked lamina, the veinlets 
usually anastomosing here and there. 


Dryopteris laserpitiifolia (Scort.) C. Chr. 

Mindanao, Cotabato Province, Guinatilan, altitude 800 meters, 
Copeland s. n., September, 1933, in wet woods. Conforms to 
Beddome's description (Suppl. 84), except for being smaller 
throughout — ^fronds about 20 cm long on 10- to 20-cm-long stipes, 
the latter brownish. Only two plants were collected — ^not 
enough for a final judgment as to eventual size. No similar 
species has been known here previously. 

DRTOPTERIS PHILIPPINENSIS <Baker) Copeland nov. comb. 

Nephrodium philippinense Baker, Ann. Bot. 5 (1891) 327. 

2V. caudiculatum J. SftL, Journ. Bot. 3 (1841) 411, nomen. 

N. extensum Hooker, Sp. Fil. 4: 72 partim, non Blume. 

Dryopteris bdsilaris C. CHR., Index (1905) 254; Christ, Philip. Journ. 

Sci. § C 2 (1907) 196. 
Nephrodium basilar e Presl, Epim. (1849) 258, nomen. 

These names are all based on Cuming 10, 8U, and SS8, as 
listed by Baker, who seems to have been the first to provide a 
diagnosis for this locally common species. Christensen rejected 

56,2 Copeland: Philippine Ferns, VIII 101 

Baker's name because of the earlier described Phegopteris phiU 
ippinensis Mett., Gymnogramme philippinense Fee, Genera 181 ; 
but this does not bar the transfer of Baker's name, because it 
has not itself been transferred to Dryopteris; also, because, I 
suppose, it is not a Dryopteris anyway, but is Heterogonium as- 
pidioides Presl, although Fee and Mettenius cite Cuming 321, 
which I have not seen, and Presl cites Cuming 295. 

The diagnoses of Baker and of Christ are alike essentially 
wrong in one respect — if our Cuming plants and theirs are 
alike : the plant is not glabrous. Cuming 84, represented by two 
sheets in hand, is very evidently pubescent on rachis, costa, and 
veins, with setulose upper surface, particularly near the margin. 
Cuming 10, the type collection, is an older specimen, and might 
be called glabrescent in anticipation, but no part is really gla- 
brous; the secondary vein, running to the sinus, is most per- 
sistently bristly. Cuming SS8 is intermediate in hairyness. Of 
the more recent collections cited by Christ, Topping 407 and 
Copeland 204 are most conspicuously hairy. Copeland 637 has 
really naked pinnae, but is not this species. Cuming 10 and 84 
are exceptionally large. It is a common fern along streams at 
minor altitudes from Cagayan, Luzon, to Zamboanga, Mindanao, 
but is not reported elsewhere. 


Dryopteris diversiloba (Presl) Christ, in Philip. Joum. Sci. § C 2 
(1907) 199. 

Besides the original collections, Cuming 51 and 102, the only 
entirely typical distributed collection I know is Bur. Sci. 23169 
McGregor, from Laguna, near the Tayabas boundary. It occurs, 
however, on Maquiling. Now, this is the type locality of D. 
acromanes Christ, t. c. 200, whence it has been distributed by 
Topping, his 7S5. The supposed differences are that the latter 
has the pinnse more enlarged toward their apices, and the sori 
confined to the lobes. I believe that they are one species. 

It is noted that Christ published the name D. dvversiloba in 
proper binomial form, but as a subspecies of D. canescens; and 
that under it he listed a variety acrostichoides (no specimens 
cited here) with two subvarieties, rhombea and lancea. Both 
of the latter seem to me best to be regarded as species. There 
is never any difficulty in distinguishing either from D. diversi^ 
loba, and, while they are sometimes found together, it is almost 
always easy to refer each individual to one or the other. 

102 The Philippine Journal of Science 1935 

DBTOPTEBIS RHOMBEA (Christ) Copeland comli. nov. Plate 6. 

Dryopteris diversUoba var4 acrostichoides subvar. rhombea Christ, 

Philip. Journ. Sci. § C 2 (1907) 200. 
Nephrodium acrostichoides J. Sm., nomen, non Michaux. 
Pronephrium acrostichoides Presl, non Dryopteris acrostichoides 0. K. 

The type is Cuming U9 in the Philippine National Herbarium, 
which Christ had on loan when he applied the name. The spe- 
cies ranges from Nueva Vizcaya (Merrill 162, cited by Christ, 
t. c. 199, as D. diversUoba) to Mindanao. It is very common 
at low to medium altitudes, and assumes bizarre forms in all 
such places. Representative distributed collections are Luzon, 
Bataan, Copeland P. P. E. 18, Leiberg 6153, Merrill 3 ISO: La- 
guna, Elmer 18205, For. Bur. 9550 Curran, Bur. Sd. 23167 Mc- 
Gregor: Tayabas, Topping 1306: Camarines Sur, Bur, Sci. 75790 
Edano. Leyte, Bur. Sd. 4175 6, Edano. Mindanao, Surigao, 
Bolster 252; Agusan, Elmer 13271, Merrill 7333, Weber 1178: 
Davao, Copeland 503, 698: Lanao, Clemens 1078: West Misamis, 
For. Bur. 4613, 4710 Meams and Hutchinson: Zamboanga, Cope- 
land 1547, 1754, 1774. 

It differs from D. diversUoba (Presl) Christ in not being deep- 
ly nor at all regularly lobed. Only in the bizarre forms there 
are some long lobes, sometimes about the middle of the pinnsB, 
more commonly nearer their apices. The sori are usually round, 
but may tend to run together in pairs. Indusia may be wanting 
on herbarium specimens, but are present, and naked,' on young 
sori. I have suspected that this might be Goniopteris asymme- 
trica F6e, as Christ once determined it for me, but have con- 
cluded that that must be a synonym of D. diversUoba; not only 
is it based on the same two collections, but it is described as 
with frondibus pinnjato-pinnatifidis. 

DRYOPTSRIS LANCEOLA (Christ) Copeland comb. nor. Plate 7. 

D. diversiloba var, acrostichoides subvar. lanceola Christ, Philip. 
Joiini. Sci. § C 2 (1907) 200. 

The type is Copeland 250, from Mount Mariveles. Other dis- 
tributed collections are Luzon, Cagayan, Weber 1550: Nueva 
Vizcaya, Bur. Sci. 14295 McGregor: Zambales, For. Bur. 5830 
Curran: Bataan (Mount Mariveles), Williams 208: Rizal, Bur. 
Sd. 986, 1807 Ramos, Topping 633, 754, 897: Laguna, Bur. 
Sd. 9714 Robinson. Negros, Whitford 1600. Mindanao, Su- 
rigao, Bolster 261, Wenzel 2617. 

8e,2 Copeland: Philippine Ferns, VIII 103 

This species is much more uniform than is D. rhombea, vary- 
ing only moderately in the degree of contraction of the fertile 
frond, and within fairly narrow specific limits in other respects. 

DRTOPTERIS MAQUILINGENSIS Capeland sp. nov. Plate 8. 

D. rhombeae afflnis pinnis majoribus haud apices versus dila- 
tatis; rhizomate repente, paleis fuscis parvis late lanceolatis acu- 
minatis vestito; stipitibus arete approximatis, stramineis setulis 
albidis pubescentibus demum glabrescentibus, frondium sterilium 
5-10 cm, f ertilium 20-30 cm altis ; f ronde pinnata, sterilis parte 
apicale 7-15 cm longa, 3-4 cm lata acuminata basi truncata, 
obscure vel grosse serrata pinnis lateralibus ca. 2-paribus 4-6 
cm longis 2-2.5 cm latis falcato-acuminatis basi subcordato- 
truncatis plerumque integris carnoso-subcoriaceis, fertilis parte 
apicale 6-9 cm longa basi 3 cm lata deinde ad apicem acutam an- 
gustata, dentata vel basin versus inciso-dentata, pinnis lateralibus 
ca. 3-paribus infimis ca. 4 cm longis vix 2 cm latis plerumque 
undulato-crenatis ; axibus minute pubescentibus ; soris primo glo- 
bosis, longitudinaliter seriatis; demum subconfluentibus, indusiis 
vestigialibus et mox evanidis. 

Luzon, Mount Maquiling, altitude 350 meters. Type in 
Philippine National Herbarium, Copeland, November 1932, nu- 
mero nondum praeditus, in Pterid. Philip. Exsic. distribuen- 
dum; ibid., Elmer 18169. 

While this is a relative of D. rhombea, I cannot bring it within 
the range of that very variable species. Beside the difference 
in form of pinnae, obvious in dried specimens, it is in nature 
distinct in texture and color, being thicker and darker. Herbaria 
which have Elmer's Maquiling plants may compare his 18205, 
which is typical D. rhombea, and 18169, which in the Philippine 
National Herbarium is sterile D. maquilingensis with very long 
stipes. Except in the degree of dimorphism, D. maquilingensis 
seems to be fairly uniform — ^for its group. 


Athyrium Elmeri Copeland, Elmer's Leaflets 2 (1908) 399; Philip. 
Journ. Sci. § C 3 (1908) 285. Described from Negros. 

To this, I refer a plant from the west base of Mount Apo, 
Guinatilan, Cotabato, altitude 900 meters, in very wet forest, 
Copeland, September, 1933, with freely and sharply incised pin- 
nules. Whether this and the Negros plant are identical or not, 

104 The Philippine Journal of Science 1935 

they are closely related, as shown by the common possession 
of dark-margined and toothed brown pale \ not noted in the 
original description. 

LOMARIOPSIS PAPYRACEA Copeland sp. nav. Plate 1, fig. 3. 

Rhizomate scandente, 7 mm crasso, apice paleis castaneis lan- 
ceolatis attenuatis 6 mm longis vestito, deorsum glabrescente ; 
frondibus plantae juvenilis simplicibus usque ad 30 cm longis, 
4 cm latis, caudatis, basi longe attenuatis, stipite 10-15 cm 
longo, plantae normalis pinnatis, stipite 10-25 cm longo, pinnis 
7-paribus vel pluribus, articulatis, subsessilibus, 15-18 cm lon- 
gis, 4 cm latis, apice abrupte caudatis cauda 10-15 mm longa, 
basi inaequilateraliter cuneatis, papyraceis, venis conspicuis; 
fronde fertile 20-30 cm longa stipite incluso, pinna apicale 6-7 
cm longa, 16 mm lata, lateralibus pauUo minoribus. 

Mindanao, Davao Penal Colony, Copeland, August 30, 1932, 
in low, wet woods. Type in Philippine National Herbarium. 
Only one fertile frond found with pinnae still present. 

Related to L. Sndthii, but distinguished by the subsessile pin- 
nae and the small fertile fronds with broad pinnse. This would 
be identified more reasonably as L. cochinchinensis, from which 
it differs in general in the same respects. That species has 
not been reported from the Philippines. It might reasonably 
be expected in Mindanao ; but I have thought it more reasonable 
to describe the plant in hand as new, than to report an exten- 
sion of range on the strength of a very aberrant specimen ; L. 
subtrifoliata, also described from Mindanao, is too different in 
texture to invite comparison. 

Hypolepis repens J. Sm., non Presl. 

Saccoloma moluccanum C. Chr., Index 372, as interpretation of H. re- 
pens J. Sm. 
Hypolepis nigrescens HooKiai, Sp. Fil. 2 : 66, as to the Luzon plant. 

The plant in question is Cuming 271, of which we have two 
ample sheets, sterile, but unquestionably D. scandens. The 
spiny axes, which must have been responsible for Smith's wrong 
identification, as well as for Hooker's, make it impossible that 
they had Ithycamlon (S. moluccanum) in hand. Dennstaedtia 
scandens is a rare species in the Philippines, collected only five 
times in the last thirty years. 

Be, 2 Copeland: PhUippine Ferns, VIII 105 

POLYPODIUM APOENSE .ropcland »p. nor. Plato 9. 

Rhizomate hrevemt gracite, inter baseos stipitum paleis paucis 
minutis f ulvis obtus o vestito ; stipitibus f asciculatis, filif ormibus, 
1-1.5 cm longis, glandulif eris ; f ronde 3-4 cm longa, 4 mm lata, 
subpinnata, tenuiter herbacea, ubique (inferne densius) pilis 
minutis clavif ormibus nitidis plerisque bicellularibus vestita; 
segmentis oblongis, ca. 1.2 mm latis, patentibus, obtusis vel suba- 
cutis, integris vel sinuatis, monophlebiis ; soris basalibus, 
approximatis et interdum trans rhachin confluentibus. 

Mindanao, Cotabato, Mount Apo, altitude 2,000 meters, Cope- 
land s. n,, ad truncos muscosos. 

Distinguished from P. sikkimense Hieron., P. pseudotricho- 
manoides Hayata, and P. pvlogense by the pubescence; in this 
respect like P. glandvloso-pihsum Brause, which by description 
should be firmer in texture and have subtriangular segments. 

POLYPODIUM PACHYCAULUM Copeland sp. hot. Plate 1(». 

Ctenopteris, rhizomate breve, 5 mm crasso, radicibus, basi- 
bus stipitum et paleis atrocastaneis 6 mm longis anguste lan- 
ceolatis attenuatis irregulariter insigniter ciliatis dense immerso ; 
stipitibus caespitosis, exarticulatis, 1-2 cm longis, rhachibusque 
atrofuscis, pilis purpureis 1 mm longis vestitis; f ronde ca. 10 
cm longa, 2-2.5 cm lata, utrinque angustata, coriacea, deorsum 
pinnata, pinnis in parte superiore decurrenti-confluentibus, 
erecto-patentibus, integris, 2 mm latis, inferne pilis minutis 
atropurpureis dense pubescentibus superne subglabris; venis 
occultis, simplicibus; soris medialibus, superficialibus, demum 
saepe confluentibus. 

Mindanao, Cotabato, Mount Apo, altitude, 2,000 meters, Cope- 
land, September 6, 1932, ad truncos muscosos. 

The aspect is that of P. nutans, not that of any other species 
known to me with pubescent surfaces. 

GRAMMITIS MICROTRICHA Copeland sp. nor. Plate 11. 

Inter G. pleiosoram n. comb. [Polypodium, Mett., Linnaea 36 
(1869) 128] et G. sumatranam: n. comb. [Polypodium, Baker, 
Journ. Bot. (1880) 214], pube stipitis 2 cm longi densa et bre- 
vissima, lamina subcoriacea minute pubescente sat dense et bre- 
vissime ciliata, ca. 12 cm longa, 10-12 mm lata, crenata, venis 
pluriramosis immersis, soris superficialibus, globosis, pluriseria- 
tis vel irregulariter sparsis. 

106 The Philippine Journal of Science laas 

Mindanao, Cotabato, Guinatilan (west base of Mount Apo), 
altitude 1,000 meters, Copeland s. n., September, 1933, on mossy 

Grammitis sumatrana is found higher on Mount Apo (de- 
scribed there as Polyp, pleiosoroides) ; it is decidedly more co- 
riaceous, less pubescent and with rather longer hairs, and mostly 
elongate sori or at least receptacles. 

LOXOGRAlfME MAJOR Copeland sp. nor. Plate 12. 

L. parallelae similis quondam eacumque confusa, statura ma- 
jore costa stipiteque ebeneis distincta, rhizomate valido paleis 
3.5-5 mm longis lanceolatis griseo-f uscis acuminatissimis integris 
vel hinc illinc minutissime spinulosis (parietibus cellularum ex- 
currentibus) vestito; stipitibus 1-5 mm inter sese distantibus, 
1-2 cm longis, validis, subapplanatis, ebeneis ; f ronde plerumque 
30 cm longa, 10-12 mm lata (rarius usque ad 50 cm longa), 
utrinque angustata, costa nigra valida utraque facie applanata; 
soris 1-4 cm longis, costae fere parellelis, baud imbricatis. 

Mindanao, Cotabato, Mount Apo, altitude 1,800 meters, Cope- 
land, s. n., September 7, 1932. 

This is the fern already known from the neighboring Mount 
Matutum and tentatively reported [Philip. Journ. Sci. 38 (1929) 
153] as L. parallela. Collecting it again on Apo, and more 
amply, I find the greater stature to be constant, and that it 
is further characterized by disproportionately stouter, black 
stipe and costa. 

The Kinabalu (Borneo) species similar in stature has been 
collected again by Mrs. Clemens, 28986. It is subsessile without 
black axes, in which respects it is like typical L. parallela. 

GONIOPHLEBniM TERRESTRE Copeland sp. nar. Plates 13 and 14. 

Schellolepis, rhizomate late repente, gracile, 2 mm crasso, 
paleis castaneis persistentibus 2-3 mm longis apicibus setaceis 
squarrosis ubique vestito; stipitibus 8-12 cm longis; f ronde 20- 
30 cm longa; pinnis utroque latere 3-5, maximis sterilibus 8 cm 
longis, 14 mm latis, valde acuminatis, basi subsessile cuneatis, 
irregulariter serrulatis, herbaceis, plerisque minoribus et an- 
gustioribus; venis seriem unam areolarum magnarum, rarius 
alteram interruptam minutarum includentibus ; soris superficia- 
libus, magnis. 

Luzon, Mount Maquiling, altitude 500 meters, Copeland, De- 
cember, 1932, ad saxas humidas. 

Like G. persicifolium in form of pinnae and superficial sori; 
distinct from this, and from Polypodium Koningsbergeri as de- 

w, 2 Copeland: Philippine Ferns, VIII 107 

scribed, in the usually single row of areolae. Abundant on one 
area of wet, sunny cliffs, to which the rhizomes are strictly 
appressed, fruiting at all seasons. Eare in the same neighbor- 
hood on mossy branches. 


Reported from the Philippines from Mount Apo only, col- 
lected there by Warburg, Elmer, and three times by myself. 
The Apo plant, as represented in our herbaria, is atypical, being 
less rigidly coriaceous and with less-immersed sori than that of 
Java; that is, it is less completely different from the probably 
ancestral P. taeniMum. 

CERATOPTERIS SILIQUOSA (Liim»ns) Capeland comb. nor. 

Acrostichum siliqtLOSum LiNN-ffiUS, Sp. Plant. (1753) 1071. 

Acrostichum thaUctroides LiNN-ffiUS, ibid, 

Ceratopteria thalictroides Brongn, [BuU. Soc. Philom. (1821) 186, 

not seen], Benedict, Bull. Torr. Bot. Club 36 (1909) 463. 
Ellobacarpus oleraceus Kaulfuss, Enum. (1824) 149. 

For the numerous other synonyms see Benedict, loc. cit., and 
Christensen's Index. I take up the most tenable specific name, 
not because there is any satisfaction in a change, which there 
emphatically is not, but because I have material of economic 
interest to publish on this plant, and do not wish in a paper 
of essentially economic interest to discuss nomenclature, nor to 
use a name subject to correction. 


[Th« photogrraphs were prepared by the Bureau of Science. The drawings were 
mostly made by Alicbusan.} 

Plate 1 

Fig. 1. Opkioglossum Ramosii Copeland. Type, X 1/2. 

2. Opkioglossum Ramosii Copeland. Spike, X 1. 

3. Lomariopsis papyracea Copeland. Pinna, X 1/2. 

Plate 2 

Cyathea bipinnatifida Copeland. Type: i, Fronds, X 2/5; 2, pinna, 
X 4/5; S, pale®, X 1-3/5. 

Plate 3 

Cyathea umbrosa Copeland. Type: i, Pinnule, X 1; ^, squamule of costa, 
X 50; ^, palea of rachis, x 50; 4, palea of stipe, x 50. 

Cyathea lepifera Copeland. Cotype: 5, Pinnule, X 1; ^, squamule of costa, 
X 50; r, palea of stipe, x 50. 

Plate 4 

Cyathea pteridioidee Copeland. Type: i, Pinna, x 1; ^, scale on costa, 
X 50. 

Plate 5 

Dryopteris squamipes Copeland. Type: i, Mounted plant, X 2/5; g, pinna, 
X 6/5; S, palea from rachis, X 18. 

Plate 6 

Dryopteris rhombea (Christ) Copeland. i, Mounted (type) plant; ^, va- 
riability of pinnae, beginning at top, Leiberg 6150, Bataan; Topping 
1S06, Tayabas; Bur, Scl 41756, Leyte; Bolster £5S, Surigao; Weber 
1178, Butuan; Copeland 698, Davao; Copeland 151^7, Zamboanga; Cope- 
land 177 U, Zamboanga; For, Bur. J^61S, Mount Malindang. All X 7/8. 

Plate 7 

Dryopteris lanceola (Christ) Copeland. Type: 1, Mounted plant, X 0.4; 
2, fertile pinna, X 3.2; S, sterile pinna, x 1.6. 

Plate 8 

Dryopteris maquilingensis Copeland. Type: 1, Mounted plant; 2, pinna, 
X 1; ^, scale of rhizome, X 25; 4, sorus, x 30, 

Plate 9 

Polypodium apoense Copeland. Type: 1, Frond, X 2; 2, segment, X 5; 
S, fragment, showing trichomes, X 20; 4, palea, X 60. 


110 The Philippine Journal of Science 

PlJlTB 10 

Polypodium packucatdum Copeland. Type: 1, Plant before mounting, 
X 2/5; J0, sterile fragment, X 4; 5, fertile fragment, X 4; -^, palea of 
candex, X 20, 

Plate 11 

Grammitis microtrieha Copeland. Type: i, Plant before mounting, X 1/2; 
$, fruiting frond, X 1; ^, fragment of same, x 2; 4, hairs on stipe, 
X 70; 5, hairs on lamina, X 70. 

Plates 12 

Loxogramme major Copeland. Type: 1, Plant before mounting, X 8/8; 
B, palea, X 20. 

Plate 13 

GoniophleMum terrestre Copeland. Type: Plant before mounting, x 1/3. 

Plate 14 

GoniophleMum terrestre Copeland. Type: i. Normal pinna, X 1; ^, very 
broad pinna, X 1; 3 and 4, palae, x 30; 5, fertile fragment, X 1. 

foi'KLWh: PHU.IPriM. I''RRK^, \*IIl 

I t'liii IP .!<>} R.\, 8rt.. ."»»>. x< 



Vm'KI,AKU: I'HliJPI'IN-R FWi\'r^, \'11|. i 

I Pmi ir. Jtints, Sri.. .%ii, \"< 


iUn'Hh.wn: Piiiupi'inf l-'nixs. VilL 

[Philip. Jciurn. Scl, 56, No. 2. 


^•r4~' ^ 

\tf\i ' • '''■•' 

i;H.- : 


riH'KLA.\'!»: PlIlMri'iNK Fkkns, Vtif. 

•nil. IF. J^il'KN. Si'!., 't*K Kf 










I'OFKI.AM*; PlIil.lPI'INK FkKNs. \{ll. 

1 Thii ir, .loTRX, v^t !., .'»»*». N» 


■■if if i%-^ 



roi'KI.ANI* : PiitlJI-plM-: I'^KKKS, \" III. ! 

! PiMijf. Jofijx, Sri,. :.!;. N< 



^or^:LANl•: Philii'I'INk Fhux.^, \ I II. I 

! rnil IP. Jori{\. i^* I., ."hI. Xr. 


ro|-K!,A.\'|k: i*lllf.Il'!'I,\K FkRNs, Vlll.l 

I'HiLii'. Jariix. Sfi.. .*,•;. Si 



C'<ii-KiaM>: I'HiiJi'Pix'F Fkkxs. \' Hi. I 

! I', JorHx. Sn.. Ttn, Xo. :!. 

... ■■"■/■ 




t'lil'l-.I.A Nii:'flNK FkR\S, X'lllJ 

I I'M 1 1. 1 P. .loriiN. Srf.. 5it. N* 

ih *%" "'t*'^"^-^ 



PLATE 10. 

t\n'KL-A\h: PiiiiJi'FiNK Fkkns. \'lll. 

I I'lULii'. .Jours, Sri., r^;, N«». 


•^-^'•^^ «# 

"I •. 

:f: 1 

PLATE 11. 

r«>rKi.AM»: Piiii.iMiXK Fkkxk. \1H.| 

I t*n 11 IP, .1 ort<x. H«"i.. "»t», N« 

^3 i 


W ^^ 

PLATE 12, 

COPEI! , Fi'jj.vs. \ III. 

! i'nti.i!». j«irKx. Hri.. r.i;. i\< 

^%. ^ 


1 1^ 

'^- W 


PLATE 13. 

r<»l'KIAX!K Plll!.Il'l'!!\|.: l^'KltNS. \'! II. i 

I Pfii! IF. Jot ns. St I., .".!;. X t 

>;' i & 

PLATE 14. 


By F. B. Serrano * 

Of the Bureau of Science, Manila 



Pineapple mealy-bug wilt is by far the most destructive mal- 
ady known to affect the pineapple plant, Ananas comosus 
(Linn.) Merr., particularly Smooth Cayenne, which is consid- 
ered the premier commercial variety. It is characterized by a 
general wilting of the plant, with or without green spotting on 
the leaves. It assumes a number of forms depending upon 
the age, the succulence, and the vigor of the plant, as well as 
the size and time of onset of the initial mealy-bug infestation. 
A large population at the onset of the initial mealy-bug infes- 
tation produces quick wilt (Plate 1, fig. 4), while a small num- 
ber of insects produces slow wilt (Plate 1, fig. 4, x). The 
younger and the more succulent and vigorous the plant is, the 
quicker it succumbs to quick wilt. 

Pineapple mealy-bug wilt is caused by the toxic secretion of the 
pineapple mealy bug, Psmdococcus brevipes (Ckll.).(4) It is 
found everywhere the Smooth Cayenne pineapple is grown, caus- 
ing almost complete collapse of the entire field, in several in- 
stances. While much has already been said by various in- 
vestigators about its destructiveness, etc., no definite informa- 
tion about its control is given in any of the available literature 
on the subject. In view of this and of its economic importance, 
attempts were made immediately after its true cause was known 
to find ways and means to control it or at least to check its 

^ This is a continuation of "Pineapple mealy-bug wilt in the Philippines," 
Philip. Journ. Sci. 55 (1934) 363-377. 

^ The writer wishes to express his thanks to the Philippine Packing Cor- 
poration for the splendid cooperation extended him in supplying practically 
all the material and labor used in conducting the experiments. He is also 
grateful to Dr. G. 0. Ocfemia, of the Agricultural College, Laguna, for 
reading the manuscript. 


112 The Philippine Journal of Science 1935 


Full knowledge of how the malady spreads is of prime im- 
portance in devising means of control. A more or less exhaus- 
tive investigation of the matter was therefore made, and it was 
ascertained that the pineapple mealy bug may be disseminated 
in three ways ; namely, through infested plant material, through 
the agency of ants, and to some extent, through its own volition. 

When plant material such as crowns or tops, slips, suckers, 
stumps, seedlings, and fruits are removed from an infested field 
without previous treatment for pineapple mealy-bug control, the 
pest may be introduced to new fields. This practice is in fact 
responsible for the rapid spread of the malady. 

Another sure way of spreading the pest is through the aid of 
two species of ants, Pheidole megacephala (Fabr.) and Solenop- 
sis geminata Fabr. var. rufa Jerdon, which feed on the honey- 
dew secreted by the mealy bug and which, in return, take care 
of the young mealy bugs. 

The pineapple mealy bug may also invade, although to a very 
limited extent, new pineapple plants through its own volition. 
It can crawl from leaf to leaf and from plant to plant. This is 
especially true with the gray strain or the green-spotting type (4) 
which does not seem to like remaining very long in one spot. 


The spraying experiments for the control of the pineapple 
mealy bug were carried out in the pineapple fields of the Phil- 
ippine Packing Corporation situated at Santa Fe, Bukidnon 
Province, Mindanao, from January to September, 1930. That 
year was climatically speaking unusually dry for that locality. 
It was, nevertheless, favorable for the mealy bugs and, fortun- 
ately, also for spraying. 


Materials and methods. — ^To find out the best insecticide for 
the control of the pineapple mealy bug the efficacy and feasibil- 
ity of soapsuds spray, nicotine-soap spray, and oil-emulsion 
spray were tested. The experiment was laid out in January, 
1930, on thirty-six double rows of standard length of 4-month-old 
Hawaiian Smooth Cayenne variety, spaced 56 by 22 by 18 inches. 
The plants were uniform in size and general vigor, with more or 
less uniform mealy-bug infestation, but showed no signs of wilt 
at the time. 

5«'2 Serrano: Pineapple Mealy-bug Wilt 113 

The block was divided into three parts of twelve rows each. 
The first part was treated by spraying with 2 per cent soap- 
suds ^ rows 1 to 3 and 7 to 9 at monthly and bimonthly intervals, 
respectively, for four months, leaving rows 4 to 6 and 10 to 12 
unsprayed as check. The second part instead was treated 
similarly with nicotine soap ^ for rows 13 to 15 and 19 to 21, 
and with oil emulsion ^ for rows 25 to 27 and 31 to 33, leaving 
the rest unsprayed as check. By means of a compressed-air 
spray pump (Plate 2) all of the plants except the check were 
sprayed one by one, each leaf thoroughly wetted, and the spray 
material driven between the closely appressed leaves and the 
heart especially, where an abundance of the mealy bugs is 

Results. — At the close of the experiment observations on the 
general health and vigor of the plants were made. The results 
are presented in Table 1. 

It is shown in Table 1, first, that the mealy-bug wilt of pine- 
apple may be checked by spraying with either soapsuds, nicotine 
soap, or oil emulsion ; second, that monthly application is more 
effective than bimonthly application ; and third, that these sprays 
are practically the same in efficiency, soapsuds having 73.9 per 
cent; nicotine soap, 77.2 per cent; and oil emulsion, 83.3 per cent 
efficacy. The wilting of some of the sprayed plants may, per- 
haps, be explained by the fact that when the spraying was 
started several of the plants had already heavy mealy-bug infes- 
tation, although no signs of wilt were visible at the time. Had 
the spraying been started earlier or before the mealy bugs got 
well established, better results might have been obtained. 

An after-effect of the treatment noticed is the scorching of the 
leaves sprayed with either nicotine soap or oil emulsion. Such 
harmful effect is, however, much more pronounced in the lat- 
ter, in many instances causing decay of the heart or terminal 

* Soap solution was prepared by dissolving 378.5 grams of laundry soap 
(Sefiorita) in water to make 19 liters of 2 per cent solution. 

* Nicotine-soap was prepared by mixing 15 liters of 2 per cent soapsuds 
with 4 liters of concentrated tobacco decoction. 

" Oil emulsion was prepared by emulsifying a mixture of 378.5 cc rope 
oil and 378.5 cc water, in which 378.5 grams brown pottery clay had been 
slaked. Emulsification is done by passing the mixture through a spray 
l5ump several times, then adding to it enough water to make 19 liters of 
2 per cent emulsion. The mixture is repassed through the pump once 
more before using. 

289853 2 


The Philippine Journal of Science 


bud of the plant. Unfortunately the injury does not become 
manifest until after a month or so. The decay of the heart in- 
duces the plant to branch off in the form of five or more suckers, 
the production of which delays the maturing of the crop. 

Table 1, — Showing comparative efficacy of soapsuds, nicotine soap, and oil 
emulsion for the control of pineapple mealy -bug wilt. 



Spraying intervals. 





1~ 3 

4~ 6 

7- 9 


Average. _ _ 







Per cent. 



Bimonthly. __ _ 






Nicotine soap 

Monthly _ 







Check.__ _... 

Nicotine soap 

Bimonthly _ 




Average- _- 

Oil emulsion 

Monthly . 





Check _ 

Oil emulsion 





- _ _ -_ 


Owing to the cheapness, handiness, availability, and especial- 
ly the efficacy of soapsuds spray, which proved to be practically 
as high as that of either nicotine soap or oil emulsion, it was de- 
cided to ascertain the best way to attain the most satisfactory 
results with it. Laboratory tests (the inclusion of which is 
deemed unnecessary) clearly showed that of the various brands 
of laundry soap, Seiiorita is the one that compares most fa- 
vorably with Chinese Yellow in efficacy. Chinese White and 
Lenox are practically the same. Because Seiiorita soap has a 
more stable and more reliable consistency than Chinese Yellow, 
it was chosen for all subsequent experiments of this series as 
well as others in which soap is the main constituent of the spray 

Materials and methods, — ^This experiment was laid out in 
April, 1930, on fifty-four double rows (56 by 22 by 18 inches) 
of 4-month-old Hawaiian Smooth Cayenne plants in the pine- 
apple fields of the Philippine Packing Corporation, at Santa Fe, 
Bukidnon, showing uniformity in general stand and vigor, as 


Serrano: Pineapple Mealy-bug Wilt 


well as in the distribution and extent of the pineapple mealy-bug 
infestation. Soapsuds in three dilutions, 1, 1.5, and 2 per cent, 
were employed in three different series ; that is, as plain soapsuds, 
as soapsuds with brown pottery clay, and as hot soapsuds. So- 
lutions of the various concentrations were prepared from a 10 
per cent stock solution. 

Starting with plain soapsuds, plants of rows 1 to 3 were 
sprayed thoroughly, one by one, with 1 per cent solution; rows 
7 to 9, with 1.5 per cent; and rows 13 to 15, with 2 per cent, 
leaving rows 4 to 6, 10 to 12, and 16 to 18 unsprayed as check. 

With the compressed-air pump spraying was done monthly for 
from two to five months, depending on the prevalence of the 
mealy bugs. The remaining rows were treated in like manner 
but with soapsuds containing 0.2 per cent brown pottery clay, 
for rows 19 to 21, 25 to 27, and 31 to 33, and plain soapsuds 
heated to from 45° to 50° C. for rows 37 to 39, 43 to 45, and 
49 to 51, while the other rows, 22 to 24, 28 to 30, 34 to 36, 40 
to 42, 46 to 48, and 52 to 54, remained unsprayed as check. 

Results, — The observations made at the close of the experiment 
are presented in Table 2. 

Table 2. — Shoming increase in the relative efficacy of soapsuds of different 
concentrations caused by the addition of clay and by heating. 









1- 3 

4- 6 

7- 9 





Soapsuds, plain 

Per cent. 








Per cent. 











Soapsuds, plain _ _ _ 




Soapsuds, plain 



Check - 

Average __ 

Soapsuds plus 0.2 per cent day 










Soapsuds plus 0.2 per cent day 




Soapsuds plus 0.2 per cent day 





Soapsuds, 45-50 °C 










Sopasuds, 45-50 °C 




Soapsuds, 45-50 0C_ 




— " 

116 The Philippine Journal of Science 1935 

Table 2 shows that the optimum concentration of soapsuds for 
the control of pineapple mealy-bug wilt seems to lie between 1.5 
and 2 per cent, and that the addition of 0.2 per cent brown 
pottery clay or heating the solution to from 45'' to 54° C. renders 
the soapsuds more effective. These facts are indicated by the 
number of applications needed by each treatment, by the num- 
ber of healthy and diseased plants, and by the average percent- 
age of efficiency. As in the first experiment it was also noted 
that the plants that came out diseased in the sprayed rows are 
those which have had more or less heavy mealy-bug infestation, 
although they showed no signs of wilt when the experiment was 

In field practice one of the things tried to make the spray 
more effective and giving satisfactory results was to brush the 
mealy bugs off with a long-handled brush (Plate 2), simulta- 
neously with spraying. High pressure was also found to contri- 
bute to the effectiveness of the spray. The higher the pressure 
at which the spray is delivered the quicker and the surer the 
mealy bugs are stripped of their woolly and waxy covering, hence 
the quicker they succumb to the treatment. 


While spraying is necessary for the control of pineapple mealy- 
bug wilt once the pest has gained foothold in the field, greater 
emphasis must be placed on using only mealy-bug-free plant 
material (suckers, slips, crowns, etc.), particularly in starting 
new plantations. This was clearly demonstrated by the check 
plants used in the infestation experiments. (4) Whenever pos- 
sible plant material should be obtained only from mealy-bug- 
free fields. If the plantations are infested, however, efforts 
should be made to select the plant material from mealy-bug- 
free individuals. Otherwise effective treatment of some kind, 
like dipping, must be resorted to before setting out the plants. 

As it has been shown conclusively in the writer's earlier ar- 
ticle (4) that pineapple mealy-bug wilt is caused by the pine- 
apple mealy bug, Pseudococciis brevipes (CklL), it is axiomatic 
that the exclusion of the insect will prevent the malady. It 
was felt to be of prime importance, therefore, to search for an 
effective means by which the plant material could be freed from 
the mealy bugs, granting that such plant material was infested. 
In the preceding spraying experiments it was revealed that 
soapsuds are just as good as, if not better than, either nico- 

56,2 Serrano: Pineapple Mealy-hug Wilt 117 

tine soap or oil emulsion for the control of the pineapple mealy 
bug, because of their handiness, cheapness, feasibility, stability, 
and high order of efficacy. The following experiments, started 
November 1, 1930, in Santa Fe, Bukidnon Province, were de- 
signed to explore the suitability of soapsuds as a dipping 


Materials and methods. — Three different concentrations of 
soapsuds were prepared; namely, 44 liters of 1 per cent, 44 
liters of 1.5 per cent, and 44 liters of 2 per cent; each solution 
was divided into four equal parts and kept in empty gasoline 
cans. Stumps of wilting 8-month-old Smooth Cayenne plants 
showing mealy bugs in abundance were collected and trimmed. 
Nine of such stumps were immersed in the first solution; three 
of them were "fished out'' after 10 minutes, three after 20 
minutes, and the remaining three after 30 minutes, and each 
batch was placed in a separate, properly labelled, enameled pan 
with the lips lined with Tanglefoot preparation to prevent ants 
from taking away any mealy bugs that might survive the treat- 
ment. The same number of stumps were treated similarly in 
the second solution, another batch in the third solution, and 
still another in the fourth, the latter two solutions being kept 
at a temperature ranging between 54° and 55° C. during the 
treatment. This completes the tests in the first series of 1 
per cent solution. Tests on the remaining two series of 1.5 
per cent and 2 per cent solutions were carried out in exactly 
the same manner and with the same number and kind of stumps 
infested with an abundance of pineapple mealy bugs. Care was 
taken to make the conditions of the experiment in all essentials 
as nearly identical as possible throughout the whole series. 

Results. — The following day the stumps of the three series 
were examined minutely, one by one, with the help of a bino- 
cular for the mealy bugs present, dead or alive. The results 
are given in Table 3. 

Table 3 shows that for the control of pineapple mealy-bug wilt 
soapsuds are more effective with pottery clay than without, and 
that their efficacy is increased by more than twice when the solu- 
tion is heated to and maintained at a temperature of 54° to 
55° C. during the treatment. It is also shown that at room 
temperature 1 per cent is as ineffective as either 1.5 per cent 
or 2 per cent; but when used at 54° to 55° C. they are equally 


The Philippine Journal of Science 




Si, O 

§ e 

CO »«2 

I i 

a g 








00 N eo o «* 00 -^ 

© © t- t- © 

©a ''jt so © iH o o» 

d lo iH d d 


eo "til 0> O ^ U9 9> 

o -^ lO o> © 


rW T-t 



O CO O O M <o o 

© CO 04 © © 

eo* <e d d ©1 CO o 

o b^ b.' d d 

^ to O O U3 CO o 

© kO © o © 

rH r-l tH 

r-l T-( T-( 



O «J O 0> M © C<l 

J§ S S? 2 S 

© 00 CO r-4 0> 


b- t- CO 00 tH N rH 
CO CO CO lO CO Tl« t- 



© CO -*!< b- •** 


O 00 © © VO © o 

© © "^ © © 


04 00 00 rH 

00 © 


-<»< ea N « 

04 01 




^ 00 O © CO lO © 

© b- OS © © 



w' & O <> ^ G G 

© rM © © © 




CO Tf © © -^i^ U3 © 

© -^ "^ © © 

r-4 r-l tH 

T-l r-J f-4 






•^ SO © •«* iH rf © 

OS 00 04 © b- 

00 iH 00 01 t- 




xl« "^ CO © '* 




»H N © © »-< «0 © 

© g M © © 



00 «* -^ eo 

s ^ 





N CO 00 © o> 00 eo 

© t- CO 01 © 



eo ud 00 d t> 00 00 

© b« © OS d 


oj CO 00 © CO eo OS 

© eo ^ OS © 


r-i rH 




04 01 © b- CO 


f-l © CO >uO r|i 



iH C« U» to 09 CM lO 

b- Ol eo © U3 


»5 © <o © ^ © © 

© W © VO © 


-^04 b- Tjf r^ 


^ -^ 

t- t- no lo t- t- wa 

lO t- t- U3 liO 

04 04 U3 iO 04 09 U3 

to 03 04 U3 to 







o © © © lo u» lo 

to © © © © 

tH 04 01 04 01 



iiiili 1 



1 44 14^ t 4» » 

4^ 1 4> t 4» 


1 U 1 O 1 O 1 

V 1 O 1 tf 

i 1 i 1 U i 

. I N {04 j 04 

fl o a <=> fl d c 

a ft ft •& 0. 3, a 

1 ii ii 

04 j 01 • 04 

<=> fl o q' d 

3l § 1 3 
•a at ft|. 

-§ -f 'S -§" ^ 

g g g g g 


a A Pi o. a 

«S 0? c4 «8 01 «1 (« 

ee c? c? S^ el 

O O O O O O O 

o o o o o 

CO CQ 03 CO CO Oa 02 

60 CO W M to 

56,2 Serrano: Pineapple Mealy-bug Wilt 119 

effective, rendering mealy-bug-infested stumps absolutely free 
from^ the infestation in ten minutes if 0.2 per cent pottery 
clay is added to the solution, and in twenty minutes if no pottery 
clay is added. 

Another point brought out by these results is that in spray- 
ing 1.5 per cent soapsuds gives 100 per cent killing power or 
efficiency, whereas preparations at even higher concentrations 
are practically useless as dipping solutions unless applied hot. In 
order, therefore, that a spraying solution of given concentration 
may be effective as a dipping solution it must be used hot. It 
is pertinent, perhaps, to mention in this connection that heat 
alone at a temperature of 54° to 55° C. does not kill the mealy 
bugs ; when infested plant specimens were immersed in tap water 
at the same temperature for the same duration, none of the 
mealy bugs was killed. 


It has been amply demonstrated in the preceding experiments 
that soapsuds of any reasonable concentration are not sufficiently 
effective as a dipping solution for the control of the pineapple 
mealy bug unless used hot. It was desirable, therefore, to find 
out the best temperature and time exposure for each type of 
pineapple plant material. 

Materials and methods. — Ninety liters of 1.5 per cent soap- 
suds were prepared and divided equally in six empty gasoline 
cans numbered from 1 to 6. Solution 1 was heated, and upon 
reaching the temperature of 39° to 40° C, at which it was main- 
tained, nine trimmed stumps of 8-month-old wilting Smooth 
Cayenne plants were immersed in it; three of them were 
"fished out" after 40 minutes, three after 50 minutes, and the 
last three after 60 minutes, and each set was placed separately 
in an enameled basin with lips lined with Tanglefoot preparation 
to prevent ants from taking away any mealy bug that might sur- 
vive the treatment. After this, three batches consisting of nine 
suckers weighing about 150 grams each, nine slips of about 100 
grams each, and nine crowns of about 50 grams each were treated 
similarly, one by one, while the temperature of the soap solution 
was maintained at 30° to 40° C. The same kinds and number of 
plant materials were treated, in an identical manner, but with 
one series at 42° to 43° C, one at 45° to 46° C, one at 48° 
to 49° C, one at 51° to 52° C, and another at 54° to 55° C. 


The Philippine Journal of Science 


Table 4. — The mimber of living and the number and percentage of dead 
mealy bugs found on each plant material and the degree of scorching of 
each kind of plant material after immersion in 1.5 per cent soap solution 
at different temperatures and exposures. 


2 I 

of solu- 



5 4-55 





Alive. Dead. Efficacy 








Per cent.' 









Suckers, ± 

150 grams 

each . 





scorch- ! 








— do..... 













Slips, ± 

100 grams 


Crowns, ± 50 
grams each. 









Sli|,htly scorched. 





Severely scorched 










' Do. 


: Do. 



Results. — The following day all of the stumps were minutely 
examined, one by one, under a binocular, and the mortality of 
the mealy bugs present on each noted. This having been com- 
pleted, all of the shoots and crowns were gone over, group by 
group, the next day, for any sign of scorching that the treat- 
ment might have produced thereon, and classified as normal, 
slightly scorched, or severely scorched, as the case might be. 
The results are given in Table 4. 

It is shown in Table 4 that a 100 per cent killing of the 
pineapple mealy bugs can be effected by soaking the plant ma- 
terial 60 minutes in soapsuds of 1.5 per cent concentration 
with 0.2 per cent brown pottery clay at a temperature of 39 "" 
to 40° C; 50 minutes at 42° to 43° C; 35 minutes at 45° to 
46° C; 20 minutes at 48° to 49° C; 15 minutes at 51° to 52° 

Snci:A\«»; I'l \f-, \ri'i,r MF\i\-Htt, W wi i 

[l*IIILIP. Jot it.\. Sci , .".«.. \, 

^'^^.- ' .^^ 





Si'.iti{A\4» : Pi\K\rii,K Mi-Ai 

*!i 11.11'. Jui'HX, Si'L. ,h;. X< 



-^', '■■■ - <'• ,ivv-,- :r^. 


56,2 Serrano: Pineapple Mealy-bug Wilt 121 

C, and 10 minutes at 54° to 55° C. It is also shown that 
suckers weighing about 150 grams each can withstand soak- 
ing for 20 minutes at 48° to 49° C. without danger of be- 
coming scorched ; slips weighing about 100 grams each, for 35 
minutes at 45° to 46° C; and crowns weighing about 50 grams 
each, for 50 minutes at 42° to 43° C. It is quite evident 
from these results that pineapple plant material of all types 
may be freed from mealy-bug infestation by soaking in hot 1.5 
per cent soap solution, in the following order: 

Stumps, 15 minutes at 51° to 52° C. 
Suckers, 150 grams, 20 minutes at 48° to 49° C. 
Slips, 100 grams, 35 minutes at 45° to 46° C. 
Crowns, 50 grams, 50 minutes at 42° to 43° C. 
Seedlings, small, 60 minutes at 39° to 40° C. 

The results of both spraying and dipping experiments as pre- 
sented in the preceding tabulations become more significant in 
the light of Carter's (1) revelation of the results of his labora- 
tory and field tests in dipping and spraying for mealy-bug control 
in Hawaii, stating among other things, that various commercial 
oil emulsions provide a satisfactory means of control, although 
the susceptibility of the plants to scorching imposes severe re- 
strictions on their use. The writer found this to be so when 
comparison between the efficiency of soapsuds, nicotine soap, 
and oil emulsion for the same purpose was made. 


Another possibility to render the plant material free from 
the mealy bugs is vacuum fumigation. That this possibility 
merits consideration is shown by the report of Hagan(3) in 
Hawaii, that vacuum fumigation with hydrocyanic acid gas at 
the rate of 10 ounces sodium cyanide to 1,000 cubic feet of space 
for one to one and one-half hours and 28 inches of vacuum at 
the start, gave satisfactory results. Similar treatment with 
either carbon bisulphide or chlorpicrin resulted, however, in 
great injury to the plant material. 


Scouting to locate the focus of infestation is of fundamental 
importance and should go hand-in-hand with spraying. In fact 
it should precede spraying. This is especially necessary when 
plants are still young and the mealy bugs not yet well estab- 
lished. Its usefulness cannot be overemphasized considering the 

122 The Philippine Journal of Science 1935 

fact that the malady does not become conspicuous until after 
the infestation has gone too far to respond to drastic and costly 
treatment. An early discovery of its occurrence is of fundamen- 
tal value. This can be effected by proper scouting under trained 
personnel at least once a month. 


The pineapple mealy bug has a number of natural enemies 
and predators, such as the brown lacewing sympherobid, Sym- 
pherobius augustus; the ladybird beetle, Cryptolaemus montrov^ 
zieri Muls.; Lobodiplosis pseudococci Felt; (2) and a grasshopper, 
Conocephalus saltator Sauss. The first two were introduced 
from Hawaii by the Philippine Packing Corporation, in 1930, in 
Bukidnon Province, Mindanao. Efforts should be made to en- 
courage the rapid multiplication and spread of such benefactors, 
for under ordinary conditions they should be able to hold the 
mealy bugs under control. 


While there are natural enemies and predators attacking the 
pineapple mealy bug, there are also on the other hand insects 
protecting it. Two species of ants, Pheidole megacephala 
(Fabr.) and Solenopsis geminata Fabr. var. rufa Jerdon are 
known to do this. They feed on the honeydew secreted by the 
mealy bug whose young they protect in return from enemies. 
Keeping these ants under check will therefore not only enhance 
the activities of the natural enemies and predators of the pine- 
apple mealy bug but will also adversely affect the general vigor 
and fecundity of the mealy bug itself. 

Controlling these ants, particularly the red variety, Solenop- 
sis geminata Fabr. var. rufa Jerdon, is not an easy matter. 
However, clean culture and surface mulching by harrowing the 
ground of the plantation, whenever possible, may help in driv- 
ing them away. Simple soapsuds spray as employed for the 
mealy bug treatment is effective and cheap enough to use for 
the black ant, Pheidole megacephala (Fabr.), which may be 
killed in its dugouts by pouring the solution into the holes. The 
red ant, on the other hand, never appears greatly concerned 
when treated with any of the three insecticides which proved 
destructive to the mealy bug. Fumigants like chlorpicrin, 
carbon bisulphide, and hydrocyanic acid gas may prove effective 
in the control of both ants, and their applicability should be care- 
fully and thoroughly studied. 

5^2 Serrano: Pineapple Mealy-bug Wilt 123 


On the assumption that new fields planted to mealy-bug-free 
seeds are adjacent to infested old fields, border planting as 
supplementary means of checking the malady seems advisable- 
This would tend to slow down the influx of the mealy bugs 
migrating from a mealy-bug-laden field, and to localize infesta- 
tion, thereby facilitating spraying operations. Border planting 
to be effective must consist of a bed of at least four rows and 
the plants must be as succulent and vigorous as those in the main 
plantation so as to provide enough inducement for the mealy bug 
to remain. These conditions are necessary for detaining the in- 
sect in the border planting and effectively checking its advance. 


1. It has been observed that the pineapple mealy-bug wilt 
may be disseminated in a number of ways; namely, through in- 
fested plant material, such as suckers, slips, crowns, etc., through 
ants, and through the mealy bug's own volition. 

2. Field tests have shown that the pineapple mealy-bug wilt 
may be controlled by spraying about once a month with 1.5 per 
cent soap solution. The addition of pottery clay (about 0.2 
per cent), preferably the brown type, renders the solution more 

3. Nicotine-soap solution and oil emulsion have proved to be 
practically as effective as soapsuds, if not more so, in killing 
the pineapple mealy bug. Owing, however, to their scorching 
effect on plants, which is quite serious, particularly in the case 
of oil emulsions, their application is considered rather precarious. 
Hence, soap solution is considered preferable. 

4. Laboratory tests have shown that of the different brands of 
laundry soap tried the Chinese Yellow has the highest efficiency, 
followed closely by the Seiiorita. The Chinese White and the 
Lenox are practically of the same strength. 

5. For dipping purposes soapsuds at any reasonable concen- 
tration are practically worthless unless used hot, as follows: 
Stumps, 15 minutes at 51° to 52"^ C; suckers, about 150 grams, 
20 minutes at 48° to 49° C; slips, about 100 grams, 35 minutes 
at 45° to 46° C; crowns, about 50 grams, 50 minutes at 42° to 
43° C. ; seedlings, 60 minutes at 39° to 40° C. Complete freedom 
from the pineapple mealy bug is secured by dipping infested 
plant material in 1.5 per cent soap solution at these temperatures 
and corresponding time exposures. One-hundred-ninety-liter 

124 The Philippine Journal of Science 

barrels or large vats equipped with heaters may be used for 
this purpose. 

6. As in spraying", the addition of pottery clay (about 0.2 per 
cent), preferably the brown type, renders the dipping solution 
more effective, while heating to 45° to 50° C. increases its effi- 
cacy more than twice. 

7. Proper scouting to locate the focus of infestation is neces- 
sary for successful spraying. It is especially needed in young 
plantations where the mealy-bug colonies may actually become 
established rapidly although no outward manifestations of the 
malady are yet shown. 

8. Quick reproduction and spread of the pineapple mealy- 
bug's natural enemies and predators — such as, Sympherohius 
augustics, Cryptolaenms montrouzieri Muls., Lobodiplosis pseu- 
dococci Felt, and Conocephalits saltator Sauss. — should be studied 
and the results obtained therefrom disseminated among pine- 
apple growers. 

9. Keeping under check the ants Pheidole megacephcda 
(Fabr.) and Solenopsis geminata Fabr. var. rufa Jerdon, which 
are in symbiotic association with the pineapple mealy bug, 
Pseudococctcs brevipes (CklL), will materially help in keeping 
the malady under control, 

10. In starting new plantations in districts where the malady 
is more or less well established border planting is a necessary 
precaution, particularly when ants are abundant. 

11. The need for using nothing but mealy-bug-free plant 
material in opening new plantations cannot be overemphasized. 
This may be accomplished by either selecting plant material 
from mealy-bug-free fields or properly treating the mealy-bug- 
infested plant material prior to planting. 


1. Cakteb, W. The use of insecticides on pineapple plants in Hawaii. 

Journ. Econ. Ent. 24 (1931) 1233-1242. 

2. Felt, E. P. A new enemy (Lobodiplosis pseudococci sp. nov.) of the 

pineapple mealy-bug [Pseudococcus brevipes (Ckll.), in Hawaii] and a 
list of gall midge enemies of mealy bugs. Journ, N. Y. Ent. Soc. 
61 (1933) 119-123. 

3. Hagan, H. R. Vacuum fumigation of pineapple planting material. 

Journ. Econ. Ent. 24 (1931) 1002-1012. 

4. Serrano, P. B. Pineapple mealy-bug wilt in the Philippines. Philip. 

Journ. Sci. 55 (1934) 363-377. 


Plate 1 

Pig. 1. Three-month-old Smooth Cayenne plantation free from wilt. Plant- 
ing material was treated with dipping soap solution before plant- 

2. Same as fig. 1, eighth month; the plant is still free from wilt. 

3. Same as figs. 1 and 2, thirteenth month; the plant is still free 

from wilt. 

4. Thirteen-month-old Smooth Cayenne plantation on the verge of 

collapse due to wilt. Plant material was not treated for mealy- 
bug control before planting. Plants in the foreground are vic- 
tims of quick wilt, while those marked x are suffering from slow 
wilt. All figures about X 0.02. (All photographs by the author.) 

Pl^te 2 

Compressed-air spray pump, with a handmade long-handled brush on the 
left side; about X 0.2. (Photograph by C. S. Angbengco.) 



By Masaaki Tokunaga 

Assistant Professor of Entomology, Kyoto Imperial University, Japan 


The remarkable dipterous insect Nymphomyia (Uba Toku- 
naga, the subject of this investigation, was first collected in 
March, 1932, by myself along a torrential stream at Kibune, 
Kyoto (200 to 420 meters in altitude), a locality famous for 
its rich insect fauna and also known as the habitat of other 
peculiar insects; such as, Epiophlebia suprestes Salys (Odo- 
nata), Deuterophlebia sp. (Diptera), Galloisiana nipponensis 
Caudell and King (Orthoptera) , Lyea^nopsis sugitanii Matsu- 
mura (Lepidoptera), etc. In December, 1932, a brief report 
on this new species, from the taxonomic viewpoint, was pub- 
lished in Annotationes Zoologicae Japonenses, the substance of 
this paper having been previously presented at the Eighth 
Annual Meeting of the Japanese Zoological Society held in Octo- 
ber, 1932, at Taihoku, Formosa. 

The present paper deals with the external morphology of this 
interesting form and gives, in addition to full descriptions of 
the structures, discussions on the homology of the sclerotization, 
concluding with a suggestion on the phylogenetic position of 
this fly among the dipterous insects. 

Most of the nymphomyiid flies examined were collected in 
the fall of 1932 at Kibune, and preserved in 70 per cent alcohol 
after fixing in Carnoy's fluid. Living and dried material, and 
also that directly preserved in 70 per cent alcohol, as well as 
fixed material, was freely used in this study. For supplemen- 
tary purposes various dipterous and other insects in the uni- 
versity collection were also examined. Dissections were carried 
out under a binocular microscope by means of steel and glass 
needles. When the material needed softening, it was treated in 
5 per cent potassium hydroxide. Ziehl's carbolic fuchsin or 

^Contribution from the entomological laboratory, Kyoto Imperial Uni- 
versity, Japan, No. 36. 


128 The Philippine Journal of Science 1935 

Mann's methyl-blue-eosin was used for the staining of the dis- 
sected pieces. Fine dissected pieces were observed under high 
magnification after mounting in glycerine-alcohol or cedar oil. 
This investigation was undertaken under the direction of 
Prof. Dr. Hachiro Yuasa. For his invaluable assistance, which 
made this study possible, I desire to express my hearty thanks. 


The morphological terminology used in this paper is mainly 
adopted from that of MacGillivray (1923) and of Tokunaga 
(1930 and 1932), supplemented from that of various modern 

In its general features this fly resembles a certain caddice- 
fly larva rather than any known dipterous insect. It is small 
and delicate, measuring about 2.3 millimeters in length, and is 
white in life. The important characteristic structures of this 
fly may be briefly described as follows : 

1. The head is prognathous in type, with a very large oc- 
cipital foramen, and elongated cephalad, forming a snoutlike 
projection. The mouth opening is located in the characteristic 
pocketlike mouth cavity and all of the mouth parts are almost 
completely atrophied. The compound eyes are contiguous on the 
ventral side. The two ocelli are located on the lateral sides of 
the head capsule and are very large. The antennse are of 
brachycerous type characterized by the presence of the inter- 
segmental sensillse. 

2. The thorax is elongated and cylindrical. The pronotum 
is completely subdivided into two, paired, lateral halves on the 
middorsal line, and the prosternum is represented by a very 
large, undivided, primary sternum to which a pair of the latero- 
sternites has probably united. The mesopostscutellum is ex- 
tremely developed. The mesopleura are small and membranous, 
and the mesosternum is very large, being divided into the pre- 
sternum and the basisterno-sternellum. The metanotum is atro- 
phied, while the metasternum is very large and represented only 
by the basisterno-sternellum, the presternum being lost. Gen- 
erally the development of the endoskeletons is very poor. 

3. The wings are very narrow, cuneiform, with very long 
marginal setae and a well-developed ambient vein. The veins 
are very few and obscure. The squama and the alula are also 
greatly reduced. The halteres are normal in structure, but 
their basal sclerites are extremely reduced. 

5^' 2 Tokunaga: A Nymphomyiid Fly 129 

4. The three pairs of legs are separated widely from each 
other, and the forelegs are articulated on the lateral side. The 
trochantin is atrophied. All the legs are provided with elongated 
coxse and trochanters, and each coxa is provided with a basi- 
costal suture. Femora and tibiae are subdivided into the prox- 
imal and distal regions by membranous rings. The claws and 
their accessory structures are simple. The empodium is slender 
and elongated. The calcanea is slender and squamous. The 
planta auxiliae and pulvilli are wanting. 

5. The differentiation of the pregenital segments is very 
slight in both sexes. The eighth abdominal segment in each 
sex is provided with special paratergal projections. The genital 
structures are exposed, and in neither sex of the concealed or 
telescopic type. The cerci of each sex are large and prominent. 
The postgenital segments are greatly reduced and uniformly 
fused with each other. The abdominal spiracles are all wanting. 


The head capsule of this fly (text fig. 1 and Plate 1, figs. 1 
to 4) is quite unique both in shape and structure. The head is 
small, conical, elongated cephalad forming a snoutlike projec- 
tion, and cephaloventrad, a labiumlike projection. The sclerites 
of the head capsule are fused with each other, forming a common 
capsule, w^hich is entirely covered with a fine pubescence and 
somewhat setigerous with slender setae. The sutures of the head 
are almost atrophied except for a shallow and incomplete suture 
on the caudodorsal region. This is a part of the epicranial stem 
(es). The occipital foramen (of) is very large, slightly smaller 
than the width of the head, round, without distinct odontoideae 
but uniformly thickened along the margin. The endoskeletons 
(tentorium) of the head (supratentoria, pretentoria, and meta- 
tentoria) and tentorinae (supratentorinae, pretentorinae, and me- 
tatentorinae) are all wanting. Thus, in this species the muscles 
of the head are directly attached to the cuticular dermis without 
the intermediation of the chitinized tendons. 

The compound eyes (ce) are subequal in size and shape in 
both sexes, oval in shape in the lateral aspect, without velutinous 
hairs on the surface, contiguous on the ventral side but widely 
separated on the dorsal. The distance between them is two- 
thirds the vertical length of an eye. The facets of the eyes 
are biconvex and granulose in appearance; there are about 
forty on each eye. The oculata (ol) is large, thickened, dark, 

289853 3 

130 The Philippine Journal of Science isss 

somewhat broader on the ventral side, and with a fine pubes- 
cence on its ectomarginal surface. On the midventral side the 
basal parts of the paired oculatse are completely fused with each 
other, but their distal marginal parts are widely separated, 
so as to appear T-shaped in cross section at this point. The 
ocelli (oc) are very large, paired, present laterally or caudad 
of the compound eyes, or on the widest region of the head 
capsule; each ocellus consists of a common hyaline exocuticula 
and a biconvex, glassily hyaline, independent cuticular lens under 
which a variable amount of pigmental material is arranged in 
masses. The dorsomedian ocellus is wanting. 

The antennae are short, 5-segmented, and brachycerous in type 
in both sexes, located very near to each other on the dorsal 
side of the base of the snoutlike projection (sn) or cephalad of 
the compound eyes, but independently inserted in the small anta- 
corise under a small common visorlike projection. This blunt 
structure (vr) is a mere projection of the head capsule and 
has no homologous relation with the ptilinum of the higher 
Diptera. The antennaria is completely fused with the common 
head capsule and not visible as an independent sclerite. The 
basal segment, scape (sp), is pyriform and entirely covered 
with a fine pubescence, with several (usually five) small, soft, 
slender setae on the distal margin. The proximal part of this 
segment is narrow, and this margin is thickly chitinized but 
without distinct antartis. The second segment, or pedicel (p), 
is somewhat smaller than the scape, and spherical; it is also 
covered with a fine pubescence on the entire surface and with 
minute hyaline sensillae on the distal half but without setae, and 
articulated to the scape by a broad coria. The third segment 
(fli), proximal segment of the flagellum, is rather more mem- 
branous than the other segments; it is very large, about 1.5 
times as long as the preceding two segments taken together, 
somewhat spoon-shaped and flattened, entirely covered with a 
very minute pubescence, with minute sensillae, but without verti- 
cils, ordinary setae. The remaining two segments of the 
flagellum are very minute and completely bare, not mem- 
branous, but thickened and slightly brown under transmitted 
light. The proximal of these two segments (fl2) is cylindrical, 
subequal in length to its width, while the terminal one (fls) is 
elongated, conical, somewhat needlelike, and three or four times 
as long as the penultimate segment. Between the distal two 
segments of the flagellum are three clavate, hyaline, interseg- 

5«. 2 Tokunaga: A Nymphomyiid Fly 131 

mental sensillse (is). These are very difficult to detect as they 
are quite hyaline. 

The mouth cavity (mc) opens very near the ventral labium- 
like projection at the innermost part of the thickened cuplike 
concavity that is located on the proximoventral side of the 
snoutlike projection of the head capsule. The mouth parts are 
represented only by a small papilliform membranous projection 
(la), which shoves a double nature, having two papillse, and 
located at the entrance of the mouth cavity or on the caudal 
side of the pocketlike concavity. This membranous projection 
of the mouth parts may be taken as a reduced labial appendage 
like the labellum (paraglossa) , judged by its location, but its 
exact morphological homology is not known. The other mouth 
parts — such as, the labrum, mandible, maxilla, maxillary palpus, 
hypopharynx, etc. — are completely atropied. 

The basipharynx and postpharynx are well retained in spite 
of the extreme reduction of the prepharynx and hypopharynx 
and the complete atrophy of the external trophic organs. The 
basipharynx (text fig. 1, bx) is a short canal extended obliquely, 
consisting of a thick dorsal wall and a thin ventral membrane, 
and invested with thin common ring muscles. In cross section 
the shape of the basipharynx is somewhat crescentic, the dorsal 
wall being convex ventrad and the ventral wall concave dorsad 
in the state of relaxation of the dilators, as usual with Diptera. 
The anterior end of the basipharynx is thinly membranous and 
directly continuous with the mouth concavity completely losing 
the hypopharynx and propharynx, whereas the posterior end is 
thickly chitinized and provided with a thicker muscular invest- 
ment. The cornu, or tuberculous projection of the caudal end 
of the basipharynx, is not developed as highly as in the Diptera 
in general. 

The oesophageal pump (text fig. 1, op) is slender, supported 
horizontally, as the head is prognathous in type, at the center of 
the head capsule, passing between the supra- and suboesophageal 
ganglia. The cephalic region of the oesophageal pump is com- 
paratively large and gradually narrowed caudad. The cephalic 
end of the pump is expanded cephalad beyond the junction with 
the basipharynx, forming a saclike structure below the frontal 
ganglion. In the cross section (text fig. 1, a to c) the dorsal 
wall of the pump is thickly chitinized, while the ventral wall is 
almost as thick as the dorsal but not chitinized and easily flexible. 
The lateral wall is very delicate, membranous, and in the state 


The Philippine Journal of Science 


of relaxation of the dilators or of contraction of the ring muscles 
the oesophageal pump is folded vertically at this lateral side as 
shown in the series of sections a to c. Although the oesophageal 
pump is more or less developed throughout the Orthorrhapha 
Diptera, as pointed out by Peterson (1916), in the present fly 
it is somewhat different in structure from the pump in the other 
nematocerous flies. Generally in the Nematocera the cross 

12 10 

Fra 1. Head of Nymphomyia alba Tokunaga. Cross sections of the basipharynx ; a, 
tiirough the attachments of ihe anterior dilators ; 6, through those of the posterior dilators 
and, c, of the posterior region before the oesophagus, bx, Basipharynx ; en, connective 
nerve; fg, frontal ganglion; oe, oesophagus; al, oculata; op, (Esophageal pump; p, pedicel; 
sbg» subcBsophageal sanglion; ad, salivary duct; sg, salivary gland; alp, salivary pump; 
ap, scape; apg, supraoesophageal ganglion. 

section of the pump is somewhat triangular and in a contracted 
state (relaxation of the dilators) is Y-shaped. Thus, in the 
Nematocera the oesophageal pump is composed of three main 
sides, while in the present species it consists of four main sides, 
as shown in the text figure. 

The salivary system (text fig. 1) consists of the paired pyri- 
form glands (sg), paired short ducts, unpaired long common 
duct (sd), and salivary pump (sip). The salivary pump rep- 

^^' ^ Tokunaga: A Nymphomyiid Fly 133 

resents the special dilated part of the common salivary duct 
and IS just caudad of the salivary aperture. In cross section 
the pump IS somewhat crescentic, consisting of the dorsal, convex 
(m relaxation of the pumping muscles), thickly chitinized wall, 
and the ventral concave membranous wall as in the basipharynx. 
In the longitudinal section as shown in the text figure there 
is a minute fold arising from the ventral wall at the posterior 
end of the pump, whereas there is no such projection at the 
anterior end and the wall at this end is continuous with the 
innermost wall of the mouth concavity. The posterior fold 
of the pump probably acts as a valve during the pumping action, 
preventing the backward movement of the saliva. The salivary 
sclerites of the hypopharynx, such as the salivas, have not been 
detected and perhaps they are wanting; if present, they are very 
much reduced and almost atrophied. The snoutlike and labium- 
like projections are merely projections of certain parts of the 
head capsule and provided only with a pubescence rather longer 
than on other parts of the head capsule; there is a pair of 
toothlike projections (t) on the distal corners of the former 
projection. They do not seem to have any morphological 
homology with the true mouth parts. 

The setae of the head are all slender but vary in length, and 
their arrangement on the head capsule is as follows: Six pairs 
of long setae arranged transversely at the middle of the head 
capsule or caudad of the compound eyes, two pairs of small 
setae on the dorsoproximal surface of the snoutlike projection 
or caudad of the blunt dorsal projection of the head capsule 
near the antennal base, four pairs (long and short, arranged 
alternately) on the meson of the dorsal surface, one pair of 
long setae cephalad of the antennal base, one pair of long setae 
on the lateral sides of the snoutlike projection, one pair on the 
distoventral part, and three pairs on the distodorsal margin of 
the snoutlike projection respectively, two pairs of long setae on 
the caudoproximal part of the labiumlike projection or cephalad 
of the compound eyes, and one pair of small setae on the caudal 
margin of the pocketlike mouth concavity or tip of the labium- 
like projection of the head capsule. Besides the above setae, 
there are two pairs of porelike sensillae on the ventromeson of 
the head capsule or caudad of the compound eyes. 

Miiscles of the head, — On account of the extreme reduction 
of the external trophic organs, the development of the muscul- 
ature of the head is very poor. There are only twelve different 

134 The Philippine Journal of Science isss 

muscles (text fig. 1), including those of the antennae. Special 
apodemes and tentoria for the attachment of the muscles are 
not developed. 

1. Abductor of the pedicel. — This is a comparatively small 
fan-shaped muscle arising from the dorsoproximal region of the 
inner surface of the scape and is attached to the dorsoproximal 
edge of the pedicel. 

2. Rotator of the pedicel. — ^This slender muscle arises from the 
mesoproximal region of the inner surface of the scape and is 
inserted in the ventroproximal edge of the pedicel. 

3. Depressor of the pedicel. — ^This small muscle is located on 
the side opposite to the abductor. It arises from the ventro- 
proximal region of the scape and ends very near the attachment 
of the rotator. 

4. Depressor of the scape. — This is a comparatively small 
muscle on the dorsal side. It arises from the inner surface 
of the vertex and ends on the dorsoproximal edge of the scape. 

5. Abductor of the scape. — This is a short muscle on the base 
of the snoutlike projection. It arises from the laterodorsal wall 
of the head capsule, extends dorsomesad, and is inserted in the 
lower edge of the lateral part of the scape. 

6. Levator of the scape. — This muscle is almost of the same 
size as the abductor and also situated on the base of the snout- 
like projection. It arises from the laterodorsal wall of the 
snoutlike projection, extends dorsocaudad, and ends on the 
ventroproximal edge of the scape. 

7. Dilators of the basipharynx. — ^These are three strong paired 
muscles, which arise from the dorsomesal wall (frontoclypeus) 
of the snoutlike projection, extend caudoventrad, and are inserted 
in the chitinized dorsal wall of the basipharynx through the 
muscular investment. 

8. Muscles of the salivary pump. — ^These consist of two pairs 
of small muscles directly arising from the ventral membrane 
of the basipharynx and are attached to the central region of 
the chitinized dorsal wall of the salivary pump. 

9. Anterior dorsal dilators of the oesophageal pump. — These 
muscles are found cephalad of the supracesophageal ganglion, 
composed of five paired bands, of which each of the middle three 
pairs consists of two partially fused muscle bands, so that there 
may have been eight pairs of muscle bands originally. These 
dilators arise from the dorsal wall of the vertex between the 
compound eyes, extend ventrad, and are inserted into the dor- 

**'^ Tokunaga: A Nymphomyiid Fly X35 

socephalic part of the oesophageal pump passing through the 
investment of the ring muscles just caudad of the cephalic 
expansion of the oesophageal pump. 

10. Postenor dorsal dilators of the oesophageal pump.— These 
are a pair of long strong muscles on the caudal region of the 
head capsule; each is composed of two partially fused muscle 
bands. Each muscle arises from the dorsomesal part of the 
occipital region, and passing under the supraoesophageal ganglion 
extends ventrocephalad to be attached to the dorsocaudal region 
of the oesophageal pump. 

11. Posterior lateral dilators of the oesophageal pump. These 

are located on the caudoventral region of the head capsule. 
Each of these muscles consists of one strong band, which arises 
from the inner surface of the postgenal region, ventrad of the 
ocellus, extends mesocephalad, and ends on the ventrocaudal 
region of the oesophageal pump very near the attachments of 
the posterior dorsal dilators. 

12. Ring muscles of the basipharynx and oesophageal pump.— 
No specially isolated and independent sphincters have been found, 
such as those in Phlehotomus, Simulium, Anopheles, and Culi" 
coides. In the case of the present fly the entire structure of the 
basipharynx and oesophageal pump is provided with a continuous 
muscular investment, which consists of a great many ring mus- 
cles. This muscular investment is especially thickened on the 
cephalic region of the oesophageal pump and thinnest on the 


The thorax (Plate 2, figs. 5 to 7) is scantily haired with 
slender setae and entirely covered with fine pubescence, extremely 
elongated, cylindrical in general appearance, subequal in dia- 
meter throughout the whole len^h, and not produced over the 
head. The cervix (cc) is broad, as wide as the head itself, 
membranous, and without the chitinized cervepisternum and 
other sclerites. The spiracles of the thorax are greatly atro- 
phied. A phragma along the caudal margin of the mesopost- 
scutellum is highly developed, but other phragmae are generally 
obscure or obsolete. 

The prothorax is small, consisting of two pairs of notal, one 
unpaired sternal, and one pair of small pleural sclerites, the 
whole structure occupying the lateral and ventral sides of the 
thoracic wall since the dorsal side is intruded deeply by the 
prolongation of the mesonotum. Of these sclerites two paired 

136 The Philippine Journal of Science i^as 

dorsolateral sclerites are lateral halves of the antepronotum and 
postpronotum, respectively, which are completely subdivided on 
the dorsomeson by the produced cephalic part of the mesonotum. 
Each half of the antepronotum (an) is closely applied to the 
cephalic margin of the mesonotum by the intermediation of a 
very narrow suturelike mesocoria (mc) and separated from 
the other half of the antepronotum of the opposite side by a 
wide membrane. This sclerite is characterized by the presence 
of several slender hairs on its dorsal region. Each half of the 
postpronotum (pn) is subtriangular, without conspicuous setse, 
and partially located between the antepronotum and mesonotum. 

The ventral side of the prothorax is represented by the large, 
broad, subsquare prosternum (pst). This sternum is not di- 
vided into the presternum (sternannum) , basisternum, and ster- 
nellum, and is provided with a pair of long, slender hairs on 
its lateral margins and a pair of blunt, thickened, triangular 
projections on the lateral side. These projections are the pro- 
sternacoilse (sct), which are derived from the subcoxal element, 
the laterosternite, for the articulation of the forelegs. On the 
broad, lateral, membranous area between the pronotum and 
the prosternum there is a very small triangular sclerite, which 
is the reduced pleuron (pp). The whole structure of the re- 
duced pleuron is modified into the coxacoila (coi) for the arti- 
culation of the foreleg. This coila is probably derived from the 
episternum alone. 

The mesothorax is extremely large, occupying almost the 
whole portion of the thorax, and especially the notum is deeply 
produced into the prothorax at the cephalic part and elongated 
into the abdomen, forming an endoskeleton (^^2) at the caudal 
part. The mesonotum is divided into three main sclerites; 
namely, the long scutoprsescutum, the small scutellum, and the 
large postscutellum. The scutoprsescutum (sps2) is elongated 
and somewhat oval, formed by the complete fusion of the two 
sclerites, the prsescutum and scutum, losing the scutal suture 
between them. At the cephalic end this sclerite is provided 
with a pair of small, incomplete, pseudosutural fovese (pv, not 
the parapsidis) . Along these fovese in various dipterous insects 
setse are usually found in variable number, but in this instance 
they are completely wanting. The supra-alar setal group (sa) 
of the scutoprsescutum is always represented by a single small 
seta in both sexes. The caudal margin of the scutoprsescutum 
is slightly intruded by the scutellum along the middorsal line 
(md) . On the lateral margins of the scutoprsescutum is a pair 

^^' ^ Tokunaga: A Nympkomyiid Fly 137 

of small blunt projections, which are the reduced medalari^ 
(mas) ; m this instance these alariae are far removed cephalad 
from the wing base and not functional. Other alari^, such as 
the cephalariaB and the caudalariae, are almost atrophied. 

The scutellum {SU2) is small, situated close to the caudal mar- 
gin of the scutopraescutum, and distinctly subdivided into the 
large median scutulis (si) and the small lateral parascutules 
(pc). The scutulis is somewhat pentagonal and provided with 
two pairs of long marginal and one pair of small median setse 
in both sexes. The endoskeletons of this sclerite are more 
developed than those of the other sclerites and consist of two 
kinds of phragmse. The cephalic endoskeleton, the paraphragma 
(pa), is formed by the infolded cephalic margin of the meso- 
scutellum, and the paired lateral endoskeletons (la) are formed 
by the ental thickenings along the sutures between the scutulis 
and parascutules. The parascutulis is small, and extended lat- 
erad, forming the scutalaria (s) for the articulation of the 
wing. From the caudolateral angles of the scutulis the paired 
spirales (si) arise, and each extends laterad along the caudal 
margin of the parascutulis. 

The postscutellum (pk) is extraordinarily developed and the 
caudal half of the thorax, including the dorsal, lateral, and the 
major parts of the ventral side, is occupied by the extremely 
expanded postscutellum. This large postscutellum is divided 
by one pair of distinct sutures into the median mesascutella 
and lateral parascutellse. The mesascutella (mg) is somewhat 
shield-shaped and widely separated from the scutellum by the 
membranous coria, and its caudal one-third (euz) is deeply in- 
folded caudoentad into the abdomen. The surface of the mesa- 
scutella is covered with a common fine pubescence, but lacks 
setse, and there is a faint inverted V-shaped transversal stripe 
(st) on its meson, although this stripe has apparently no mor- 
phological significance. Each parascutella (PS2) is broad, ex- 
tended laterad and then ventromesad, directly demarcated by 
the pleurotaxis on its dorsocephalic margin, connected with the 
sternum by means of a narrow sternal membrane, and its ce- 
phalic prolongation completely separates the preceding epimeron 
(em2) into the notepimeron (712) and sternepimeron (862) by 
a deep intrusion. The caudal margin of the parascutella is 
infolded entad and forms a narrow phragma along the round 
caudal margin. On the ventral side along each mesal margin 
of the parascutella is a small semichitinized sclerite, which is 
tentatively designated the parasternite (pt) of the mesothorax. 

138 The Philippine Journal of Science lass 

The pleural region is much narrowed by the extreme exten- 
sion of the parascutella and the major parts are moreover com- 
pletely reduced to membranous areas, so that the sutures are 
all obscure. The episternum (e^2) and epimeron (em2) are 
divided by the incomplete, slightly undulated, pleural suture 
{V2) y which starts from the dorsal point of the pleuron, extends 
between the episternum and notepimeron, around the cephalic 
prolongation of the parascutella, and ends on the ventral mar- 
gin of the pleuron near the mesocoxa fossa (C2) . The episternum 
is only chitinized on its ventral marginal area and its dorsal 
notepisternal region, and the episternal suture is completely 
wanting. The chitinized notepisternal region (Plate 4, fig. 13, 
7is) is very small and closely associated with the notepimeron 
(t^2) and the pleuralifera (pr) , forming a projection for the ar- 
ticulation of the wing. The membranous sternepisteral area 
is provided with several indefinite folds, which are not true 
sutures. There is a large membranous area between the post- 
pronotum and the mesoscutoprsescutum or cephalad of the meso- 
sternepistemum. The epimeral region {em2) is widely sep- 
arated into the notepimeron and sternepimeron by the cephalic 
prolongation of the parascutella (^§2), as already mentioned. 
The notepimeron (%) is thickly chitinized, small, oval, and 
closely associated with the independent sclerite, the pleuralifera, 
and the notepisternal area on its dorsal margin and ventro- 
cephalic margin, respectively. The sternepimeron (8^2) is lo- 
cated between the parascutella and the sternal membrane. The 
coxacoila of the mesopleuron is completely wanting in this 

The sternum is located on the broad membranous area, sep- 
arated from the prosternum by a mesocoria at the cephalic end, 
and very widely separated from the metasternum by a broad 
metacoria. The mesosternum itself is divided into two sclerites, 
the presternum (prs2) and the basisternosternellum (6S2 and 
sm2) by a distinct, membranous, secondary, transternal suture 
(ss). The former sclerite is cordiform, as large as the pro- 
sternum, clearly demarked, and thickly chitinized. Along the 
caudal margin of this sclerite is a narrow secondary phragma. 
On the membranous, secondary, transternal suture is a pair of 
small setse; these setae are definitely present in both sexes, 
although they are very difficult to detect as they are very 
slender. The basisternosternellum is elongated, and its demar- 
cation is very obscure as it is reduced to a membrane but is 

^^' ^ Tokunaga: A Nymphomyiid Fly 139 

more or less chitinized on its cephalic and mesal parts. The 
cephahc part of this sclerite is provided with a pair of tiny 
thickened sternacoite (sc,) on its lateral sides and a small mem- 
branous triangular incision (fts^) on its meson. This small area 
is the basisternum proper, and the remaining large area is the 
sternellum proper. The midventral suture (ms) is clearly 
marked throughout this sclerite and more or less thickened 
entad, and the endoskeleton is very weakly developed along 
this suture. 

The metathorax is very much reduced and represented only 
by certain parts of the pleuron and the chitinized sternum. The 
metanotum is completely obsolete. The pleuron is reduced to 
a membrane, and only a small, slightly thickened area (ems) 
is found close to the abdomen or caudoventrad of the base of 
the haltere (h) . The sternum is represented by a large, clearly 
demarcated sternellum (snis). This sclerite is similar in struc- 
ture to the mesosternellum fundamentally, although it is larger, 
broader, and more thickly chitinized than the latter, being 
provided with a pair of tiny sternacoilse (sc^) on its cephalo- 
lateral corners, the midventral suture (ms) on its meson, and 
a pair of slender setae 2 near the coilse. The endoskeleton of 
the sternum is represented only by a narrow phragma along 
the caudal margin of the sternellum. The metabasisternum is 
completely reduced into a common membrane continuous with 
the ventral membranous region of the preceding thoracic seg- 


The wings (Plate 4, fig. 12) are snowy white in life, and at 
rest they are held dorsad, side by side, perpendicular to the 
long axis of the insect. They are extremely peculiar in struc- 
ture, large for the size of the body, very much elongated, fully 
as long as the body or longer, and about 5.7 times as long 
as the widest portion of the wing itself, being about 2 to 2.5 
millimeters, triangular, entirely without the anal lobe, and there- 
fore the three margins [costal (cm), anal (am), and outer 
(om)2 of the wing are almost straight, the ratio between them 
being 188 : 115 : 83. The distal angle of the wing in the male is 
slightly sharper than in the female. There are no macrotrichiae 
on any part of the surface, not even on the veins, but the entire 
margin of the wing, except for the proximal part, is thickly 

' These setae had not been detected when the previous taxonomic report 
was made (1932). 

140 The Philippine Journal of Science 1935 

fringed with very long delicate hairs, and the whole membrane 
is covered with minute microtrichise. 

These marginal hairs are all milky white in reflected light, 
but pale brown under transmitted light. Those on the costal 
margin are somewhat stronger in structure and sparser in ar- 
rangement than those on the caudal margin and about half 
as long, and gradually decrease in length distad on the distal 
quarter of the costal margin; they are arranged in two lines 
alternately; those on one line are erect cephalad, while those 
on the other line are more or less bent distad. The fringe on 
the caudal margin (anal and outer margins) is very long, being 
fully 1.5 times as long as the widest portion of the wing itself, 
and most thickly arranged on the middle part of this margin; 
most of the hairs are arranged in a single line, being erect 
caudad, but the minority of the hairs on the outer margin is 
somewhat bent proximad, and at this part the fringe very grad- 
ually decreases in length distad. 

The venation of this insect is very much reduced, being rep- 
resented only by a few veins of slight thickening, most of them 
located on the proximal quarter of the wing area and ending 
on the costal margin. Vein C is broad and pale brown under 
transmitted light, the ambient vein {av) extends along the entire 
caudal margin of the wing and is as distinct as in the costal 
margin and about half as broad as the costa. Very slight thick- 
enings of the membranes faintly indicate some of the other 
veins. The proximal broad part of the costal margin shows 
a double nature, being provided with a shallow longitudinal 
furrow; this double nature may be assumed to have been in- 
duced by the incomplete coalescence of the two veins C and Sc 
on this part. The most distinct is Ri and the stem of R com- 
bined which terminates on C at about the proximal quarter of 
the length of the costal margin and is provided with about two 
placoid sensillae {ps) on the proximal end of R. The other 
vein, terminating on C a little beyond the end of Ri, is Rs, which 
is almost atrophied on its proximal half, although faintly con- 
nected with the very slightly thickened small Rs, which is located 
at the middle of the distinct vein R. Cu is rather obvious, 
extending along the anal margin of the wing, nearly parallel 
to it, and almost atrophied before the wing margin. The other 
vein, probably M, is very faint and creaselike, located independ- 
ently or freely but in the definite position on the middle part 
of the wing without connection with the other veins, being 

5^' 2 Tokunaga: A Nymphomyiid Fly 141 

completely atrophied both on its proximal and distal terminal 


The legs (Plate 3, figs. 8 to 10) are uniformly pale yellow, 
very delicate in structure, slender, but comparatively short for 
the body length, very scantily haired with slender set«, similar 
in structure and subequal in length, except that the forelegs 
are slightly longer and the hind legs shorter than the others. 
The three pairs of legs are very widely separated from each 
other and the distances between the fore and middle legs and 
the middle and hind legs are about equal to each other, being 
sflightly more than the greatest width (on the scutellar region) 
of the thorax; the forelegs are articulated far dorsad of the 
usual ventral position, while two pairs are articulated in the 
normal ventral position. 

The coxae (ex) are unusually elongated, neither cylindrical 
nor conical, but somewhat flattened, and subdivided into the 
small proximal (p) and large distal (d) portions; the fore pair 
is somewhat stouter than the others. Each coxa is provided 
with several slender and a few curly hairs on the distal end 
and one to three slender setse on the middle part. The chitinized 
proximal subdivision is very narrow, quite cylindrical, distinctly 
bounded by a narrow, ringlike, membranous, basicostal suture 
(mc) from the main division of the coxa, provided with two 
groups (one each on the cephalic and caudal side) of micro- 
scopic erect setae; the proximal subdivision of the fore coxa is 
far shorter than that of the others. The distal margins of the 
coxse are more or less incised triangularly on the dorsal sides; 
this incision of the fore pair is larger and deeper, and that of 
the hind pair is smaller and shallower than that of the middle 

The trochanters (t) are elongated, articulated to the coxae 
with the distinct trochacoriae (to), quite cylindrical, and pro- 
vided with two to four slender setae, and two isolated and one 
double placoid sensillae (sn) on the distal part. Both ends of 
the trochanter are more or less chitinized; the proximal chi- 
tinization is larger than the distal, being provided with small 
trochantes (ta) against the trochacoilse (tc) and the distal 
margin is very narrowly chitinized along the femasuture (fs). 
The anterior two trochanters are about half as long as the cor- 
responding coxae, while the hind trochanter is slightly shorter 


The Philippine Journal of Science 


than the hind coxa. The fore trochanter is shortest and stoutest, 
while the hind trochanter is longest and slenderest, being 1.3 
times as long as the former. 

The femora (/) are largest among the leg segments, more or 
less flattened and swollen on the distal part, subdivided into 
three parts; namely, the small proximal part (p), the mem- 
branous intermediate part (m/), and the large distal (d) part, 
being immovably articulated to the trochanters. The proximal 
part is very small and quite bare except for a common fine pu- 
bescence, and at this point the femur is distinctly narrowed, 
especially on the fore and hind legs. The intermediate mem- 
branous part is broad, ring-shaped, fully twice as large as the 
proximal part, quite smooth, hyaline, and provided with several 
placoid sensillse (sn) on its ventral side. This part of the femur 
is slightly flexible. The distal part is large and fully twice as 
long as the two proximal parts taken together, with five slender 
marginal setsB, besides long setss (four each on the anterior 
two femora and one on the hind) ; its distal margin is distinctly 
thickened, forming a pair of tibiacoilse (tbc) (one on each lateral 
side) and sharply incised V-shaped on the ventral side. The fore 
femur is shortest and somewhat stouter than the others, and 
the hind femur is longest and slenderest, but about twice as 
broad as the trochanter. The fore femur is about 1.5 times 
as long as, and each posterior femur about twice or a little 
more than twice as long as, the corresponding trochanter; or 
the anterior two femora are a little shorter than, and the hind 
femur is subequal to, the preceding two segments (coxa and 
trochanter) taken together. 

Table 1. — Nymphonvyia cUba Tokunaga; relative lengths of the leg seg- 

[69 units = 0.2 mm.] 














Fore - - 














The tibise (tb) are elongated, subequal in length to the cor- 
responding femora, cylindrical and slender, except those of the 
middle legs which are somewhat swollen on the distal parts, and 
each tibia is subdivided into three parts (proximal, membra- 

^^' ^ Tokunaga: A Nymphomyiid Fly 2.43 

nous, and distal) like the femora. No tibia shows any trace of 
an apical spur or any spurlike projection. The articulation 
between the femur and tibia is strong, being provided with a 
thickly chitinized coila (tbc) and artis (tba) on each side, but 
rather simpler than the hypothetical articulation of the insect 
leg in losing such structures as the tibiaflexis, fematroclia, and 
patella. The proximal parts (p) of all pairs are small, without 
setae, and subequal in diameter; those of the anterior two pairs 
are somewhat curved ventrad and each is about one-fifth as 
long as the tibia itself, while those of the posterior pair are 
more or less elongated, nearly straight, and about one-third 
as long as the hind tibiae. The membranous parts (mt) are 
subequal in length, quite smooth, hyaline, and considerably en- 
larged on the flexor ventral surface; those on the anterior two 
legs are subequal in length to the proximal parts, respectively, 
and slightly greater in diameter; while on the hind legs these 
parts are subequal in diameter to the proximal parts and only 
half as long. The tibiae are very slightly flexible and usually 
bent ventrad at these membranous parts, as in the femora. The 
distal parts (d) of the three pairs are considerably different 
in shape and length, but similar in structure to each other; 
they are each provided with one to three placoid sensillae (sn) 
on the proximodorsal surface, three or four slender setae scat- 
tered on the proximal half, and four or five slender setae on the 
distal end. Concerning the relative length and size, this part 
of the foreleg is slenderest and longest among the three tibiae, 
being very slightly greater in diameter than in the proximal 
part and about 2.5 times as long as the proximal two parts taken 
together. The distal part of the middle tibia is large and stout, 
being more than twice as broad as the proximal part and nearly 
twice as large as the distal part of other tibiae, but far shorter 
than that of the fore tibia and subequal in size to the distal part 
of the hind femur. The distal part of the hind tibia is quite 
cylindrical, as in the fore tibia, and subequal in diameter to the 
latter, but far shorter and shortest among those of the three 
tibiae, being only about half as long as the fore tibia itself. 

The tarsi (ts) are all five-segmented, slender, and very scantily 
haired with a few minute setae only on the dorsodistal end of 
each tarsal segment. Each of the third and fourth segments 
on all the legs is provided with an unpaired microscopic spine 
(tp) on the ventrodistal end. The articulations of the tarsal 
segments are more or less oblique, and there are a rather ex- 
tensive membrane on the flexor surface and a tarsaflexis and 

144 The Philippine Journal of Science 1935 

tarsatroclia on the extensor dorsal side between most of the 
segments. Each tarsal segment is somewhat enlarged on the 
distal part, the ventral surface of which is thinner in structure 
than the dorsal. The basal and ultimate tarsal segments on 
the same legs are nearly equal in length, and the proximal four 
segments are gradually shortened distad. The fifth tarsal seg- 
ment, especially on its distal part, on all the legs is somewhat 
more pubescent with long microtrichia than the other parts 
and provided with the ordinary terminal structures but shows 
no trace of the pulvilli. The ratio in the length of the three 
tarsi is about 103 : 93 : 66, and thus the long forelegs possess 
long tarsi and the short hind legs have short tarsi, but the propor- 
tion in the length of each tarsus and the leg to which it belongs 
is quite different in the different pairs. On the foreleg the 
first tarsal segment is about half as long as the tibia and scarcely 
as long as the two following segments together. The first tarsal 
segment of the middle leg is also about half as long as the 
tibia but shorter than the following two segments taken to- 
gether; the ultimate segment is subequal in length to the first 
tarsal segment and fully as long as the two preceding segments 
together. The tarsal segments of the hind leg are slightly 
stouter than those of the other legs ; their articulations are dis- 
tinctly oblique and each segment is far shorter than the cor- 
responding segments of the other legs. Each of the first and 
ultimate segments of this leg is slightly longer than one-third 
of the tibia, or only slightly longer than half of the correspond- 
ing segments of the foreleg. Among the remaining three tarsal 
segments the third and the fourth are nearly equal in length, 
the total length of the above two segments being somewhat 
greater than the ultimate segment, and each segment is slightly 
shorter than the second tarsal segment. 

The claws (c) are symmetrical in structure, slender and 
sharply pointed, quite glabrous, unserrated, not quite smooth, 
but with very fine oblique striation on the lateral side. The 
empodium (e) is elongated and slender, slightly shorter than 
the claws, and provided with many simple hairs. The calcanea 
{en) is also slender, finely squamous, and without the onychium. 
The last two structures are directly fused with each other, losing 
such structures as the planta and auxilia. The tubercula {tu) 
is not distinctly projected distad between the claws, but thick- 
ened and emarginated ventrad, forming the *'Glenkh6cker" {gh) 
against which the claws are articulated. 

^' ^ Tokunaga: A Nymphomyiid Fly 145 


The articulations of the legs are all very simple, and all the 
coxafossse are very small. The forelegs are articulated laterally 
and widely separated from each other; each foreleg is pro- 
vided with the coxartis, which articulates against the coxacoila, 
but is without the distinct sternartis. The middle legs are 
articulated ventrally, each with the stemacoila, and the distance 
between the middle legs is least among the three pairs. The 
hind legs are also articulated ventrally with the sternartis against 
the sternacoila, but without the coxartis and the coxacoila as 
in the middle legs. The distances between the three pairs are 
subequal and very long, due to the elongation of the thorax 
(Plate 2, figs. 5 and 7). 

The articulation of the wing is very complex (Plate 2, fig. 5; 
Plate 4, fig. 13). There are four aliferae on the pleurotaxis 
for the articulation of the wing; namely, the prealifera, meda- 
lifera, subalifera (pleuralifera), and postalifera. The micra- 
lifera is completely wanting. The prealifera (pal) is small, 
thickly chitinized, and located on the dorsal part of the broad 
lateral membrane. The medalifera (ml) is oval, slightly chi- 
tinized, and located between the prealifera and the notepister- 
num, and its cephalic part is attached to the prealifera. The 
subalifera (pr) is very thickly chitinized, somewhat C-shaped, 
and forms a strong main projection for the wing base, being 
closely associated with the dorsal projection of the notepisternum 
(np) and the notepimeron (Uz) . The postalifera (pol) is oval, 
thickly chitinized, and located closely near the cephalodorsal 
margin of the parascutella (psz), and the spiralis (si) is ex- 
tended along the cephalodorsal margin of this sclerite. 

Dorsad of the prealifera or near the costalis there is a small 
spinous tubercle which is the mesotegula (tez). Besides these 
sclerites, along the lateral margin of the scutoprsescutum (spsz) , 
there is a narrow, rod-shaped sclerite which is the ponta (po) . 
The ponta forms a small, irregularly shaped structure at its 
caudal end which is complexly articulated to the prealifera, the 
subalifera, and the costalis. 

The pteralise (Plate 4, fig. 13) are reduced in number, being 
represented only by the following distinct sclerites : Terminalia 
(duritae), costalis, radialis, and analis (venellse). Such duritae 
as the sigmoidea, submedia, and navicula, and such funditse as 
the costalla, mediella, and anella are almost completely atrophied. 

289858 i 

146 The Philippine Journal of Science 1935 

The terminalia (*) is located between the subalifera (pr) and 
the postalifera (pol) and closely associated with the analis (as). 
The costalis (cs) is largest among the three venellse, and the 
radialis (ra) is very small, somewhat lunate in shape, and 
anastomosed into the costa on its distal part. These two venellae 
are articulated directly to the pleuralifera and the caudal end 
of the ponta. 

The notal processes for the articulation of the wings are 
greatly reduced. The medalaria is blunt and minute, far sep- 
arated cephalad from the wing base, and has no direct me- 
chanical relation with the wing movement. The scutalaria is 
also minute and situated close to the costalis. The caudalaria 
is obscure. The other alarise are all completely atrophied. 

The articulation of the base of the haltere (Plate 3, fig. 11) 
is very simple. The spherical proximal end, the scabellum 
(sc) of the petiole (ptl), is directly articulated to a swelling 
of the rotaxis (bsr). This basal swelling is homologous with 
the region where the various basal sclerites, such as the f unditse, 
venellse, and perhaps the duritse, may be found in the generalized 
wing base. At the cephalic region of the rotaxis (r) is a blunt 
membranous projection which is covered with distinguishable 
pubescence. This projection is the remnant of the metategula 
(tes) . The pteralise, the alariae, and the spiralis are obscure. 


There are almost no sexual differences in the structure of the 
abdomen excepting the ultimate and penultimate segments. The 
abdomen of both sexes is pale yellow, feebly chitinized, slender, 
extremely elongated, and at rest slightly curved ventrad ; super- 
ficially with nine distinguishable segments, including the hypo- 
pygium, quite cylindrical, the diameter nearly uniform through- 
out the whole length, covered with a fine microscopic pubescence, 
scantily haired with slender set® and with a few scattered 
transparent patterns, brown tubercles, and various sensillae on 
each segment. There is no trace of spiracles (text fig. 2). 

The segments of this insect, except the first and two terminal 
segments, as a rule, are not subdivided into intrasegmental an- 
nulets, but provided with cuticular structures, as follows: The 
transversal thickening, which serves as an "antecosta" (ac) 
of the intersegmental muscles and becomes hidden in the inter- 
segmental fold on contraction of the muscles, is located across 
the tergum closely along the cephalic margin, narrowed and 
highly chitinized on its mesal part but somewhat broad and thin 


Tokunaga: A Nymphomyiid Fly 


on its lateral parts. On the sternum there is no corresponding 
thickening. The distinct tubercules (tb) are arranged trans- 
versally on the caudal region of the tergum; the number of the 
tubercules is very irregular, varying from one to five pairs, and 
they are less in number on the cephalic abdominal segments 
than on the caudal; and, moreover, the male is provided with 
more of them than the female. Very close to the above tuber- 


Y X 

Fig. 2. Diagrammatic arrangement of the cuticular structures of the abdominal segment. 
ac, Antecosta; It, lateral tubercle; th tubercule; tp, transparent pattern; X-X tergopara. 
tergal line; Y-Y, tergopleural line. 

cules are three pairs of long, slender, brown setse and two pairs 
of minute sensilte. The transparent patterns (tp) are scattered 
on the tergum; two pairs of them, which are round, are on the 
middle region, and four pairs located on the cephalic region 
or caudad of the cephalic thickening are somewhat oval, 
the lateromarginal pair being more elongated. Between the 
two groups of the transparent patterns are about six pairs of 
sensilte; the cephalic three pairs are placoid, and the caudal 
three are trichoid. 

148 The Philippine Journal of Science i985 

On the lateral side, where the paratergal sclerites are located 
in certain insects, is a large, somewhat elongated tubercle (It) 
on the cephalic region where the spiracle would be expected. 
This tubercle may be derived from the modified peritreme or 
other paratergal elements. Two slender setae, one long and the 
other short, are located on the middle region of the lateral side 
where there are several minute placoid sensillse. A large, elon- 
gated, transparent pattern is found ventrad of these setae. 

The pleural region, ventrad of the dorsopleural groove (Y-Y), 
is uniformly membranous with the sternum completely losing 
the pleural sclerites. 

The sternum is not provided with tubercles and tubercules. 
There are about three pairs of slender setae in the caudal region 
corresponding to those on the tergum. The cephalic region of 
the sternum is provided with two pairs of small setae, of which 
the anterior pair is very small, hyaline, and may be trichoid 
sensillae; the others are located near each other on the meson 
of the sternum. The transparent patterns of the sternum are 
large; a very large V-shaped pattern is located on the meson, 
a united or paired pattern on the mesocephalic margin, and a 
paired one on the laterocephalic margin. 

The first abdominal segment is somewhat more elongated than 
the others, subdivided into two annulets on the dorsal side, bear- 
ing several pairs of strong setae on the dorsocephalic margin, 
usually two pairs of slender setae and about one pair of tuber- 
cules on the dorsocaudal region, two small setae on each lateral 
side, and a large V-shaped transparent pattern on the ventral. 
The various other structures of the cuticule as shown on the 
other segments are very obscure or wanting. 


The terminal two segments (Plate 4, figs. 14 to 16) are very 
different in structure from the preceding segments, and the 
sexual differences are shown most clearly on these segments. 

On the eighth, penultimate, segment of the female the tergum 
(ot) is broad, uniformly chitinized, thickly pubescent, and pro- 
vided only with three pairs of slender setae on the caudal region, 
one pair of minute trichoid sensillae on the cephalic region, and 
two pairs of transparent round or oval patterns on the cephalo- 
lateral corners; the cephalomarginal thickening is very feebly 
developed, and the placoid sensillse are quite obscure. The 
sternum (os) is almost membranous and covered with rather 
long pubescence; the cephalomesal region of the sternum is 

^^' 2 Tokunaga: A Nymphomyiid Fly 149 

swollen ventrad, while the caudomesal region is suddenly de- 
pressed dorsad. The cephalic margin (at) of the sternum is 
chitinized broadly and usually hidden in the intersegmental fold. 
Along this margin is a thin antecostal lamella. Along the mid- 
ventral line the swollen region is folded or doubled up deeply 
entad, forming a large but thin, chitinized, lamellalike endo- 
skeleton (ec). These two endoskeletons serve as attachments 
of the various genital muscles. On the caudal half of the ster- 
num or on the lateral margin of the depressed region there is 
one pair of distinct long setse accompanied by three or four 
pairs of minute trichoid sensillae very near each other. The 
caudal margin of the eighth sternum is folded dorsocephalad, 
forming a large membranous genital chamber (gc) between 
this segment and the ninth. In the innermost part of the 
genital chamber is the common aperture of the sexual glands, 
guarded by a slightly chitinized lunulate endapophysis (ento- 
gonapophysis or paramere). The lateral sides or the para- 
tergal region (pt) of the eighth segment is quite characteristic 
in structure. The whole structure is more or less thickened and 
hangs down from the lateral margin of the tergum proper along 
the lateral side of the segment, forming a longitudinal tergo- 
pleural fold. The caudal part of this fold (pj) is extremely 
elongated ventromesad along the caudal margin of the sternum 
to form a cover of the genital chamber, and is provided with a 
conspicuous long seta on its basis. Besides this flattened caudal 
projection there are two distinct tubercles (t) on the tergo- 
pleural fold; one situated on the cephalic margin of this fold 
is thickened and large, while the other is somewhat small, pro- 
vided with a minute seta, and located on the ventral margin 
of the fold or just caudad of the former tubercle. 

The terminal region, including the ninth segment, caudad of 
the eighth segment, is small, about half as long as the preceding ^ 
segment, and quarter-spherical in shape. The dorsal half of 
the spherical surface, the ninth segmacoria (sc) , is very broad, 
membranous, and without setae. The ventral half is more or 
less thickened, without tubercles and setse, but with a pair of 
large cylindrical projections on the ventrocephalolateral corners. 
This projection, the cercus (ce), is not articulated but firmly 
fused to the body wall, bears many setae, and with the nonfunc- 
tional remnant of the special muscles. At the basis of the cerci, 
on the membranous sternum and along the entrance to the genital 
chamber, is a pair of small, blade-shaped, and setigerous scler- 
ites (cc), which are the cercariae. The ventral side is mainly 

150 The Philippine Journal of Science issi 

membranous, except for the lateral margins (ep)f which are a 
part of the tergum. On this part are four very long setse, one 
of which is projected caudad, while the other three are directed 
ventrad. The ventral membranous area is rugous and sub- 
divided into two regions, the cephalic and the caudal ; the cepha- 
lic region {pp and sa) is somewhat swollen ventrad, while the 
caudal (su) is flattened. The anus (a) is located between these 
two regions. 


The penultimate and ultimate abdominal segments of the 
male (Plate 5, figs. 17 to 19) are highly modified and differ 
greatly in structure from those of the female. The eighth, or 
penultimate, segment is somewhat smaller than the preceding 
seventh, and not provided with the tubercules on the dorsocaudal 
region but with paired transparent patterns on the ventral 
surface. The cephalic margin (at) of the eighth tergum (ot) 
is narrowly and thickly chitinized and provided with a pair 
of small transparent patterns on the lateral parts of this chitin- 
ization. Besides the above small patterns there is a pair of 
rather large transparent patterns on the cephalic region of the 
tergum or caudad of the cephalic chitinization. On the middle 
part of the tergum are two small trichoid sensillae, and on the 
caudal region three pairs of long slender setse are transversally 
arranged. The caudal margin of the eighth segment is very 
thinly membranous and without microtrichia. The sternum 
(os) is somewhat thinner than the tergum and more scantily 
covered with rather long pubescence. On the cephalic region 
of the sternum are large, transparent, paired patterns and small 
trichoid sensillae, and on the caudal region is one pair of long 
setae. The lateral paratergal region (pt) of the eighth segment 
is provided with two, large, peculiar processes, which are sub- 
equal in length and parallelly extend caudad as far as the 
caudal margin of the eighth segment. The dorsal (pjt) of these 
two projections is entirely covered with fine microtrichia and 
provided with a long seta on its ventroproximal part and with 
two small thickened tubercles (t), which are quite bare and 
situated close to each other on the ventroproximal base. The 
ventral projection (ph) is slenderer than the dorsal, quite bare, 
uniformly and thickly chitinized, and provided with a hooklike 
point at its distal tip. On the ventroproximal base of this pro- 
jection is a large, smooth, earlike thickening (i), which is some* 
what thin, convex laterad, and concave mesad. The above two 

^^' 2 Tokunaga: A Nymphomyiid Fly 151 

(dorsal and ventral) large projections are not independently 
movable, but work in union, being firmly fused to the common 
thickening by the common tergosternal muscles. At rest these 
two projections are closely situated along the body wall, but 
when the common thickening is pulled ventrad by the contraction 
of the tergosternal muscles, they can be extended horizontally 
or nearly so. The lateral surface of the eighth segment, besides 
the above distinct processes, is provided with two long setse 
on the blunt colliculus and a small trichoid sensilla on the mid- 
lateral region. On the ventral side, between the eighth and ninth 
segments, is a very large membranous sedeagus (aa). The 
sedeagus is quite smooth, without pubescence, directly exposed 
externally, and supported by a chitinized framework (pa), the 
ectapophysis (gonapophysis or paramere), which subdivides the 
membranous region into two parts. The proximal region (pb) 
is very broad and as wide as the eighth segment itself, while 
the distal region (p) is very small and blunt. 

The ultimate segment consists of the ninth and other pos- 
terior segments, paired large cerci, and coxites. This fused 
segment is very small and about a quarter as large as the pre- 
ceding eighth segment; its dorsum is rather thinly membranous 
and covered with delicate pubescence; its venter is very thin 
and very scantily pubescent, and the macrotrichia is quite 
wanting on both surfaces. The cerci (ce) and coxites (ex) are 
firmly fused to this segment, and the demarcation between them 
is almost wanting. The cerci alone are distinctly demarcated 
only on the ventral aspect by the presence of the cercarisB (cc). 
The cercus is located dorsad of the coxite, extended ventrocau- 
dad, setigerous on its dorsal side, and very scantily haired on 
its ventral side (pp), hairs being represented only by two long 
setse, one on the proximomesal part and the other on the middle 
part. The basis of the cercus is very broad with a very long 
seta on its dorsomesal margin, while the distal part is sharply 
pointed and with many short setae. The cercaria (cc) is very 
slender but distinct, claw-shaped, somewhat thickened, and quite 
bare. The coxite ^ (ex) is very conspicuous, broad at its base, 
occupying the greater portion of the lateral aspect of the ninth 
segment, but slender on its distal half, extending caudad as 
far as the cercus, and gradually curved ventromesad. The en- 
tire surface of the coxite is covered with common pubescence, 

* The coxite is homologous with the subcoxa of the leg-bearing segment or 
with the limb base of the stylus and is commonly known as the "side piece." 

152 Tfie Philippine Journal of Science issi 

but without any kind of macrotrichia. Its ventral ridge is pro- 
vided with two rows of thick recurved pubescence along its 
entire length. The style * (s) is very small and slender, some- 
what curved laterodorsad and directed cephalodorsad. It is 
covered with common pubescence and without macrotrichia, 
not chitinized, and without chitinized hooks or claws. At the 
articulation of the style is a blunt lobe (/i), which is also without 
setae and covered with common microtrichia. This swelling is 
homologous with the so-called "harpe" of various nematocerous 
Diptera. The periproct (sa and su) is closely fused to the para- 
proct on the ventral side and situated at the caudal end of the 
abdomen between the cerci. 


Although the head of this fly is quite peculiar in the structure 
of all of its parts, its most extraordinary features are the pro- 
longation of the head capsule, the complete reduction of the 
mouth parts, and the lateral position of the large ocelli. 

In the Diptera the elongation of the head is of two types, 
representing two independent lines of evolution. The first type 
is derived from the modification of the mouth parts; the theca, 
which is homologous with the labial stipula of the generalized 
insect, is the main structure participating in the realization 
of this type. This sclerite often seems to take part in the 
formation of the head capsule, being located on the opposite side 
of the frontoclypeus and associated with the stipes. In various 
Diptera, such as the Culicidae (Psorophora) , Asilidse, Cyrtidae 
(Eulonchus), Conopidse, Muscidse (Glossina and Stomoxys), 
TachinidsB (Archytas), etc., the elongation of the head is solely 
due to the elongation of the theca and accessory membrane. 
In various insects with extremely long probosces, such as the 
Culicidae (Culicinae), Tipulidae {Geranomyia and Elephanto- 
myia), Empidae (Empis), Bombyliidse (Bombylius and Exo- 
prosopa), etc., the exceptional length is accounted for by the 
presence of the elongated paraglossae, which, together with 
the elongated thecae, augment the length of the proboscis. The 
second type of the elongation of the head is derived from the 
modification of the sclerites of the head capsule themselves, 
and this tjrpe is very rare among the dipterous insects, although 

* The style is homologous with the stylus of the Thysanura and is com- 
monly known as the **clasper." 

5«' 2 Tokunaga: A Nymphomyiid Fly 153 

it is quite prevalent among certain other orders, such as the 
Coleoptera (Rhynchophora), Hemiptera, Anoplura, Mecoptera, 
etc. Among Diptera this type is only shown in such nemato- 
cerous insects as the Blepharoceridse (Bibiocephala) , Culicidae 
(Psorophora) , Dixidse (Dixa) , etc., and rarely, if at all, among 
the higher Diptera, such as Brachycera and Athericera. In 
these cases the prolongation of the head is mainly due to the 
elongation and modification of the frontoclypeus. 

The head of the present insect, Nymphomyia, evidently belongs 
to the latter type, although it shows many important differences 
from the other insects included in the same type. Generally 
the elongation of the head in the second type is due to the ce- 
phaloventral prolongation of the frontoclypeus, which is more 
or less distinctly demarcated from the other sclerites, and the 
mouth parts become located eventually at the cephalic end of 
this prolongation. Moreover, the large caudal region, which 
consists of the vertex and compound eyes, usually retains a 
spherical shape or is slightly flattened and round. In the pres- 
ent insect the elongation of the head capsule is, contrary to the 
above, due to the modification of the whole head capsule. The 
cephalic snoutlike projection (Plate 1, figs. 1 to 3, sn), which 
is not provided with the mouth opening at its extremity, is not 
demarcated from the common head capsule; but, judging from 
the attachment of the muscles (dilators) of the basipharynx 
and the position of the antennae, it may be homologous with 
the frontoclypeus or some parts of the frontoclypeus. The head 
of the female Bibio (Bibionidae) is more or less modified on its 
cranial region, being slightly elongated and flattened, but the 
region cephalad of the antennae, the frontoclypeus, is very small. 
In this respect the head of Nymphomyia is quite different in 
its constitution from that of Bibio. So far as the elongation 
of the head capsule is concerned, more intimate affinities with 
the present insect may be found among the most primitive 
groups of Nematocera, such as the Tipulidae. Among the Ti- 
pulidae the head of Tipula is distinctly elongated cephalad, form- 
ing a cylindrical prolongation, which, in turn, bears a distinct 
snoutlike projection on its dorsocephalic end. This snoutlike 
projection of Tipula is quite independent in structure of the 
mouth parts and is derived from the frontoclypeus. The mouth 
parts are located on the cephalic part of the prolongation, just 
beneath the snoutlike projection. The peculiar cephalic elonga- 
tion of the head of Nymphomyia is easily derivable from that 

154 The Philippine Journal of Science isss 

of Tipula if one assumes enlargement of the part of the snout- 
like projection of the latter. A peculiar snoutlike structure 
of Nemotelus (Stratiomyiidse) also may be homologous with 
that of Nymphomyia, but the relative position of its mouth parts 
is quite different since it is hypognathous in type. 

The position of the occipital foramen has the most intimate 
correlation with the shape of the head capsule. The head, 
which is round or spherical as shown in the higher Diptera,*^ 
is almost always the hypognathous type (s. lat., including the 
orthognathous type) ; that is, vertical with the mouth parts 
directed ventrad, and the occipital foramen is located on the 
caudal aspect. The nematocerous groups are also hypognathous 
(orthognathous, s. str.), as a rule, but show gradual modifica- 
tion from the hypognathous to the prognathous in varying de- 
gree, according to the tendency of prolongation in the head 
itself. In insects provided with more or less elongated heads, 
such as Rhyphus ( Anisopidse) , Mycetophila (Mycetophilidse), 
and Bibiocephala (Blepharoceridae), the position of the occipital 
foramen is more or less excentric, having migrated dorsad, and 
more advanced modification is shown in such flies as the female 
in Psorophora (Culicidse), the female in Bibio (Bibionidae) , etc., 
which have more-elongated heads. In the extreme case the head 
exhibits a prognathous type, rather than the hypognathous type, 
or nearly so, as in Dixa (Dixidse), Limnobia and Tipula (Tipu- 
lidse), etc. In this respect such a protonematocerous group as 
the Tipulidae shows a certain similarity to Nymphomyia, but the 
superficial similarity of these two types of the head is realized 
independently in various families and genera that have no 
direct phylogenetic relations between them. It is noteworthy, 
however, that this tendency is seen only among the Nematocera 
and not among the Brachycera and Athericera. The above dis- 
cussion of the head of the present insect indicates that Nympho- 
myia should probably be included in the nematocerous group, 
although more information is needed to locate its exact phylo- 
genetic position. 

The relative size of the occipital foramen is more or less con- 
stant throughout the dipterous orders, as shown by Peterson, 
and relatively small on account of the small cervix. Even in 

■In certain parasitic groups of Diptera, such as the Hippoboscidae and 
the Streblidas, the head is highly modified secondarily from the hypogna- 
thous type, and the extreme case of this modification is shown best in the 
prognathous head of Ascodipteron. 


Tokunaga: A Nymphomyiid Fly 155 

the largest, as in the Psychodinse, the area of the foramen is 
far less than a quarter of the caudal surface of the head cap- 
sule, while in Nymphomyia the foramen (Plate 1, fig. 3, of) is 
extraordinarily large, being nearly as large as the caudal aspect 
of the head itself. Moreover, the foramen of the Diptera is 
always provided with a pair of odontoidese for the articulation of 
the cervepisterna, while in the present insect these processes are 
completely wanting, and the margin (0) of the foramen is quite 
smooth. Judging from the structures of the occiput in addition 
to other parts of the head capsule, the head of this insect shows 
a closer resemblance to the head capsule of certain eucephalous 
nematocerous larvae, such as those of the Chironomidae, Dixidse, 
Psychodidse, etc., than to the imaginal head capsules, as is evident 
in the following points: Prognathous in type, large occipital 
foramen, atrophy of the odontoideae and the tentoriae, and com- 
plete fusion of the head sclerites. In point of movability the 
head of Nymphomyia is thought to be limited within a narrower 
range than the heads of the other Diptera, which are provided 
with a small occipital foramen and well-developed odontoidese, 
since the large occipital foramen of the Nymphomyiidae is con- 
nected with the thorax by a very narrow cervacoria. The com- 
plete atrophy of the cervical sclerites supports this assumption. 

The mouth parts of the dipterous insects usually undergo re- 
duction in the number of the constituent appendages, as exhibited 
in the hypothetical case, and become more or less membranous 
as a whole. This reduction and modification occurs in various 
families and genera of Diptera in different degrees. Atrophy 
takes place in the mandible, lacinia, and glossa as the first 
step of reduction, and then in the galea as the second step. 
Both or either of the paraglossae and maxillary palpi remain 
to the last. 

Generally the paraglossae are extremely well developed among 
the Diptera and constitute the important mouth parts. Even in 
certain Chironomidae, such as the genera Pontomyia and Clunio 
which have highly reduced mouth parts, the female is provided 
with a trace of the paraglossae and the male with tiny but more 
prominent paraglossae, as I reported in 1932. So far as I know, 
except in the case of the parasitic Diptera, the complete atrophy 
of the paraglossae is only known in the Deuterophlebiidae, in 
which neither sex has a trace of the maxillary palpi or the 
paraglossae, and in the present family, Nymphomyiidae. The 
maxillary palpi are not reduced and usually have three to five 

156 The Philippine Journal of Science 1938 

segments among the Nematocera, while in the Brachycera and 
the Athericera they are highly reduced or modified and show a 
nonsegmental condition; but the reduction of the palpus is 
not very rare even among the Nematocera, and the nonsegmented 
palpi are often found among certain gall midges and true 
midges. The complete reduction, as seen in the present insect, 
however, is very rare in free-living nematocerous insects, and 
examples are found only in the famous D enter ophlebia and the 
female in Pontomyia. As mentioned above, the mouth parts of 
Nymphomyia are closely related to those of Deuterophlebia, not 
only in the degree of reduction but also in the peculiar location 
of the mouth-opening, which is in the pocketlike chitinized 
mouth concavity. This cavity (Plate 1, figs. 1 and 2, mc) is 
located on the proximoventral side of the cephalic frontoclypeal 
projection (sn), which in the latter genus is far smaller and 
blunter than in the former. The labiumlike projection (Ip) of 
the present insect is quite unique for the Diptera, and the exact 
homology of this projection is not yet known, although it might 
have been derived from such sclerites as the mentum or sub- 
mentum or, more probably, from the theca or fused stipites, 
since the latter sclerites are far more stable in the Diptera than 
the former sclerites. 

In spite of the complete reduction of the external mouth parts, 
the propharynx and the hypopharynx, it seems that the struc- 
tures of the basipharynx and the oesophageal pump and those 
of the salivary system retained (text fig. 1) suggest the pos- 
sibility of a functional action of these organs for taking liquid 
food. This question is very problematic in the natural life of 
Nymphomyia, because the epithelial layer of the midintestine 
is completely degenerated, being now represented only by a 
single delicate membrane, which has lost the characteristic epi- 
dermal structures ; the thin epithelial cells of the salivary glands 
are highly reduced, now showing no trace of secretory features 
(although often the saliva is preserved within the glands of 
the imago, it may be a remnant of the secretion produced in 
earlier stadia) ; and experimentally the fly never takes food 
offered to it, such as water drops, honey water, sugar water, 
salt water, etc. 

The presence or absence of the ocelli is vastly important 
for the phylogenetic consideration of the dipterous insects since 
the distribution of ocelli is strictly characteristic of different 
phylogenetic lines. All Athericera have three ocelli (except 
only the Conopidse, in which the ocelli are wanting) . Brachy- 

5«» 2 Tokunaga: A Nymphomyiid Fly 157 

cera also have three ocelli (except only the Tabaninae, in which 
the ocelli are usually wanting), and in the pupiparous group 
both the ocelli and compound eyes are reduced in various degrees. 
Among the Nematocera the ocelli are present only in certain 
superfamilies or families, such as the Trichoceridse, Blepharo- 
ceridae, Mycetophilidse, Anisopidse, certain ceratopogonids,® ceci- 
domyiid Lestremiinae, and Bibionoidea. The ocelli (Plate 1, 
figs. 1 to 3, oc) of the present insect are highly peculiar and 
obscure their true nature, suggesting at first the spiracles or 
the metatentorinae. The peculiarity is shown, first, in the com- 
plete atrophy of the middorsal ocellus and, secondly, in the 
lateral position of the extremely large paired ocelli. That these 
lateral organs of the Nymphomyiidae are true ocelli is shown 
by the presence of the lenslike, very slightly biconvex, glassy 
structure; of the subcuticular pigments, which are regularly 
arranged in certain limited cells; and in the absence of any 
trace of tracheal branches and other spiracular structures as- 
sociated with them that might suggest homology with the 
spiracles, as well as in the absence of thickened invaginations 
that might suggest homology with the tentorinae. Symphyla 
alone is known to possess spiracles on the head, and no other 
insect is known to simulate this peculiar location of the first 
pair of spiracles on the head. The metatentorinae are often 
shown on the caudal aspect of the head capsule among insects, 
and sometimes they become very large, as in the Tipulidae (Ti- 
pula, Ctenacroscelis, etc.). From these facts, taken in con- 
nection with the other features, these lateral organs may well 
be considered as ocelli or eyespots. This hypothesis is conclu- 
sively supported by the fact that these organs are in direct 
connection with the strong but short nerve fibers (ocellar nerve) 
arising from the protocerebral lobe. 

There are many examples among the Hemiptera, such as the 
Henicocephalidae, the Reduviidae, etc., where the median ocellus 
is completely atrophied, and only two ocelli remain conspicu- 
ous, but among the Diptera such cases are very rare, occurring 
only in certain Mycetophila (Mycetophilidae), Canthyloscelis 
(Scatopsidae), Oncodes (Cyrtidae), and Culicoides (Ceratopo- 
gonidae). Although in the hemipterous insects the two ocelli 
have migrated more caudad or caudolaterad than in the other 
orders, thq ocelli in the Diptera are almost always on the dor- 

• Jobling, B., BuU. Ent. Res. 18 (1928) 211-236, reported two vestigial 
oceUi in Culicoides. 

158 The Philippine Journal of Science i9S5 

somedian region between the compound eyes in the dichoptic 
type, and just eaudad of the eyes in the holoptic type. In both 
cases the ocelli are never as far remote from each other as in 
the present insect. Besides the above differences of the ocelli 
between this peculiar insect and all the other dipterous insects, 
the ocelli in Nymphomyia are extraordinarily large, and their 
structure as a whole is not so compact and solid as in the so- 
called ocelli. In this respect, as well as in the other features 
mentioned above, these lateral organs' of the present insect ap- 
pear of somewhat different origin from the ocelli found in the 
generalized dipterous head, although they have some affinity to 
the ocelli or eyespots on the massive head of the larvse of many 
eucephalous Nematocera, or those found on the adult head of 
certain Nematocera, such as the marine Chironomidse. In these 
cases the eyespots are represented by organized masses of pig- 
ment-filled cells, which are almost nonspecialized hypodermal 
cells communicating with poorly developed nerve endings, as 
in the larvae of Ceratopogon, shown by Hesse, where no corneal 
layer, eye lens, or vitreous layer is differentiated differing 
from the highly developed primitive ocelli. The hypo- 
dermal pigmental masses of the adult head, as shown in the 
Chironomidse, are never located in the position of the primary 
ocelli of the Diptera, but are usually found on the lateral aspect, 
as in the larval eyespots mentioned by Saunders in Paraclunio 
and by Tokunaga in Clunio and Telmatogeton. In the female 
Pontomyia these pigmental spots have migrated far ventro- 
cephalad, being situated on the cephalic aspect of the flattened 
head or ventrad of the compound eyes, and each occupies a fairly 
large area. The anomalous position of these pigmental mass- 
es, in addition to their structures, suggests a certain relationship 
between the lateral organs of Nymphomyia and the eyespots of 
the above-mentioned insects. Although Saunders suggests that 
the pigmental spots on the head of Paraclunio are some organs 
other than the ocelli, since they are not accompanied by the 
definite nerve, since the whole of them are subcutaneous, and 
since the hypodermis passing over them is unbroken, it is very 
difficult to compare such primitive organs as the eyespots with 
the ocelli found on the adult heads* or on the heads of nymphs 
and naiads. Moreover, the nerve system communicating with 
the eyespots is not always so distinct as in the true ocelli, and 
it probably is represented by very fine neurofibrillae, as in the 
peripheral nerve system, which are almost impossible to de- 
monstrate by the usual histological technic. As previously noted, 

^^' 2 Tokunaga: A Nymphomyiid Fly 159 

the so-called "ocelli" of the Nymphomyia are thought to be dif- 
ferent in origin from the primary or dorsal ocelli of insects, 
and may be more closely related to the eyespots or larval adaptive 
ocelli of the nematocerous larvse or those of certain adult 

The antennae of the present insect (Plate 1, fig. 4) are quite 
brachycerous in type in their general appearance and in the 
reduction of the flagellum segments. The antennae of the Dip- 
tera, as a whole, show a wide range of development, but in the 
majority of the genera the main line of specialization is toward 
the reduction of the flagellum segmentation. On the other hand, 
the scape and the pedicel undergo only a slight change. Ex- 
amples of the simple reduction in the number of the flagellum 
segments are found even among the nematocerous Diptera, such 
as the females of certain genera of the Chironomidae and both 
sexes of the Deuterophlebiidae, but in all these cases, although 
the length or size is more or less divergent, the differentiation 
among the flagellum segments is very slightly developed, and 
all the segments of the antennae are subequal in shape and 
structure as shown in Tanytarsus, Chironomus, Pentapedilum, 
Deuterophlebia, etc. The segmental differentiation of the fla- 
gellum of the Nymphomyiidae shows rather a closer similarity 
to that of the Brachycera, especially the Empidae, than to that 
of the Nematocera, but this similarity of the antennae is not 
thought to represent a direct phylogenetic affinity between the 
present family and the Brachycera. On the other hand, the 
presence of the intersegmental sensillae (Plate 1, fig. 4, is) on 
the flagellum in the present insect is quite peculiar among the 
adult antennae of the Diptera, since the various sensillae found 
are always located on the segments themselves and never on the 
intersegmental coriae, although the intersegmental sensillae are 
very common in the larval antennae. As mentioned above, it is 
noteworthy that the Nymphomyiidae, which show many primitive 
nematocerous characters, are provided with peculiar antennae 
whose appearance is strongly suggestive of the antennae of the 
higher Brachycera, although they are not absolutely identical 
with the latter. 

The antennarise of the Diptera are more or less closely united 
with the common head capsule, and their demarcations generally 
become obscure, both in the Orthorrhapha and the Cyclorrhapha. 
Their independent existence is plainly shown only in certain 
limited families or genera of the nematocerous groups that are 

160 The Philippine Journal of Science 1935 

provided with large antacorisB or accessory membranous areas; 
such as, the Tipulidae, Chironomidse, Trichoceridae, Culicidae, 
Mycetobiidae, etc. Both the absence of the independent anten- 
narise and the great reduction of the membranous areas of the 
antecorisB of the present fly are closely related to the modifica- 
tion of the antennae themselves. In the majority of the cases 
where the antennarise are superficially atrophied, the antacoilse, 
which properly belong to the antennarise, are more or less dis- 
tinctly retained on the membranes of the antennal sockets, and 
especially often in the Nematocera along the margins of these 
membranes, ring-shaped sclerites, which are homologous with 
the antennariae, are more or less distinctly demarcated from the 
head capsule, as I have indicated in the case of a certain crane 
fly. Thus, the antennariae, although not developed as free in- 
dependent sclerites, may be more widely present, in the state 
where they are fused with the adjacent sclerites of the head 
capsule, among the dipterous insects than Peterson and others 
suspect. In the present insect also the antennarise are more 
probably fused with the common head capsule than completely 

The sexual variation of the compound eyes is most prevalent 
among the Nematocera and Brachycera, but in Nymphomyia 
the sexes are very similar in structure, size, and shape. The 
compound eyes of the Diptera, as a rule, are either separated 
from each other, the dichoptic type, or two sides of them are 
contiguous on the dorsal side, the holoptic type, but in Nym- 
phomyia the two sides are only contiguous on the ventral side 
and widely separated on the dorsal side. The extent of the 
holoptic condition is said by Peterson to depend upon the size 
of the compound eyes and the location of the antef ossse. In spite 
of the facts that the antennae are located close to each other and 
far remote cephalad from the compound eyes, and that the an- 
tefossae are exceedingly small, the compound eyes in both sexes 
of this insect are expanded ventromesad and exhibit a unique 
modification of the holoptic type quite unknown in the Diptera. 

The visorlike ridge (Plate 1, fig. 2, vp) near and over the 
bases of the antennae is quite different in origin from the ptili- 
num,^ which is constantly found along the frontal suture of the 
higher dipterous group Schizophora, since there is no invagina- 

' In the mature adult stage the ptilinum, as a rule, is completely infolded 
entad, but very often in certain anthomyiid flies this structure remains 
external and thickened, forming a crownlike projection. 

'*' ^ Tokunaga: A Nymphomyiid Fly 161 

tion of any kind or endoskeleton associated with it. It is ap- 
parently a simple projection of the vertex, which is found in 
varying degree among the Nematocera; for example, the "frontal 
tubercle" of the Chironomidae, the "frontal crest" of the Tipu- 
lidse, etc. 

The tentorium undergoes a considerable reduction in the dif- 
ferent families, genera, or species and is associated with the 
reduction of the mouth parts. The complete degeneration of 
the tentorium occurs on the supratentorium alone, and the other 
tentoria (anterior and posterior arms) are more or less retained 
even in the head, which has no functional mouth parts, as seen 
in Chironomus, Tanytarsus, and certain other nematocerous 
flies. In a more reduced or more specialized case the tentorium 
is not developed as the endoskeleton; in this case it is repre- 
sented by the external thickening of the gene, as shown by 
Peterson in the head of Schizophora. But in the case of Nym- 
phomyia the tentorium has completely disappeared, as in the 
eucephalous head capsules of the dipterous larvae greatly differ- 
ing from the cases mentioned above. 


The peculiar feature of the thorax of this insect is mainly due 
to the extreme elongation of the cylindrical thorax. The thorax 
of the nematocerous insects, as a rule, is highly convex in both 
dorsal and ventral directions, so that the height of the thorax 
is rather greater than, or at least subequal to, the longitudinal 
length ; and the thorax of the higher dipterous groups, Brachy- 
cera and Athericera, is flattened on the dorsum and somewhat 
angulated along the lateral margins, so that the width of the 
thorax more or less approaches its length. The thorax of the 
present insect is very much elongated, being about thrice as 
long as the greatest width ; such extreme elongation of the thorax, 
so far as I know, is quite new among the dipterous insects, even 
among the lower groups, such as the Tipuloidea and Psycho- 
doidea. This elongated and cylindrical feature, which is ex- 
hibited only in the immature stages in dipterous insects, may 
be taken as a more primitive character from the evolutional 
point of view. If so, this species, which in this feature is unique 
among known Diptera, may represent a very primitive form. 
Although this unique structure of the thorax reminds one of 
that of certain immature forms rather than that of the imaginal 
forms, other morphological features, such as the possession of 

289858 5 

162 The Philippine Journal of Science 1935 

the large mesothorax, which occupies the major part of the entire 
thorax, the extremely developed mesopostscutellum and the 
highly reduced metathoracic notum, taken in connection with 
other structures, show that it is unmistakably a dipterous adult 
in form. Moreover, thoracic characters, such as the fairly well- 
developed pronotum (an and pn), although it is small for the 
extreme development of the mesonotum, and the nonangulated 
mesopleural suture (P2) show that this fly belongs to the Or- 
thorrhapha, more particularly to the Nematocera (Plate 2, fig. 
5). Finally, the large, isolated, sternal sclerites, taken in con- 
nection with their constitution and the small vestigial pleural 
sclerites, are thought to symbolize a rudimental or primitive 
condition of their own phylogenetic development among the 
dipterous insects. 

The cervix of the Diptera, as a rule, is very small for the 
size of the head capsule, while that of Nymphomyia is large 
in comparison with the small head capsule, which is subequal in 
diameter to the thorax itself. This large diameter of the cervix 
is distinctly proportional to the large occipital foramen. The 
pronotum is usually divided transversely by a secondary suture 
into two parts; namely, the antepronotum and postpronotum, 
but these parts (at least the antepronotum) are not completely 
separated longitudinally at the middorsal part. In cases where 
they are superficially separated into paired lateral portions, the 
lateral pairs are actually continuous with each other by a very 
narrow ental chitinization along the cephalic margin of the 
mesoprsescutum, as I have suggested in the case of Pontomyia. 
In the thorax of Nymphomyia, on the contrary, both the ante- 
pronotum and the postpronotum are completely and widely 
separated into paired lateral portions by a wide membranous 
area (Plate 2, fig. 6, mc) at the dorsomeson, and the chitinized 
pronotal bridge is completely lost. Moreover, the paired lateral 
portions (Plate 2, fig. 5, pn) migrate far caudad as if they 
belonged to the mesopleuron proper. This anomalous position 
of these sclerites, however, is foreshadowed by the caudal migra- 
tion of the postpronotum along the lateral margins of the 
mesonotum, as in the Anisopodidae, Mycetophilidse, Blepharoce- 
ridae, Chironomidae, Culicidse, etc. 

Another peculiar feature of the prothorax is the position of 
the coxae. Among the dipterous insects the procoxae are always 
articulated ventrally, even in the chironomid midges, where the 
prolegs are more widely separated from each other than in 

5»» 2 Tokunaga: A Nymphomyiid Fly 163 

other insects. In the present insect they are widely separated 
and articulated laterally. As a consequence the pleural area 
(Plate 2, fig. 6, pp) is very much reduced between the large 
notum and the dorsally migrated coxafossa, and is represented 
only by the procoxacoila. This remaining chitinization is prob- 
ably derived from the proepisternum alone, completely losing 
the epimeral chitinization, since among the dipterous insects 
the undivided condition of the pleuron is never found as well 
marked as in the other higher orders, and when the reduction 
occurs on the pleuron the proepisternum remains more stable 
than the proepimeron. In contrast with the very small pleura, 
the sternal region (Plate 2, fi^. 7, pst) is very large, being rep- 
resented by a large chitinized sclerite. Such a large sternum® 
is never found among the dipterous flies, and, taken in connec- 
tion with the undivided sternum, which is completely isolated 
by a membranous area from the adjacent sclerites, reminds one 
of the primary sternites (Snodgrass) of the Apterygota or the 
immature forms of the Trichoptera, Odonata, and other orders 
that have the nymphal or campodeiform early stages. 

On the mesonotum of the present insect the scutal suture 
(suture between the prsescutum and the scutum) is completely 
obsolete, but this is not rare among the Diptera, since it is 
present only in certain families, such as the Tipulidse, Tricho- 
ceridae, Tanyderidae, and Blepharoceridae, and completely atro- 
phied or greatly reduced in the majority of the Nematocera. 
The scutopraescutum (praescutum-scutum) and the scutellum are 
very similar in structure to those of the hypothetical thorax, 
except for the extraordinary elongation of the former and the 
distinct flattening of the latter sclerite. 

The extraordinary development of the postscutellum (Plate 
2, figs. 5 to 7, ph) of Nymphomyia, which is larger than the 
preceding notal sclerites isps2 and SU2) taken together, repre- 
sents the most peculiar structure of this thorax. Generally this 
sclerite is well developed in the lower Orthorrhapha, especially 
in the Nematocera, but far smaller than half of the scutoprae- 
scutum taken alone, even where it is most elongated, as in the 
Ptychopteridae. Furthermore, this sclerite is more or less ex- 

* The broad sternal sclerites are found in the thorax of the Streblid© 
and allied families, which are of the most aberrant groups of the Diptera, 
but in this case the broad sternal condition is evidently derived from the 
secondary modification of the depression of the dipterous thorax, due to 
the parasitic adaptation. 

164 The Philippine Journal of Science i^s^ 

tended ventrad or caudoventrad along the precostal margin of 
the first abdominal segment, hence the thorax, as a whole, ap- 
pears to be truncated at the caudal end of the scutellum, as 
distinctly shown in the Cyclorrhapha. This tendency of the 
postscutellum is already shown in the lower Diptera, being 
especially pronounced in such nematocerous groups as the Pty- 
chopteridse {Ptychoptera and Bittacomorpha) ^ MycetophilidjB 
{Bolitophila) , and Trichoceridae (Diazosma), while in the pres- 
ent insect, contrary to the above general character of the post- 
scutellum, this sclerite (ph and en2) is extended far caudad, and 
almost straight, deeply intruding into the first abdominal seg- 
ment ® and forming a phragma.^^ 

The postscutellum (Plate 2, fig. 5) of this insect is also dis- 
tinctly divided into paired lateral portions (parascutellse or 
pleurotergites, PS2) and the unpaired median portion (mesascu- 
tella or mediotergite, 1712) as in the generalized dipterous thorax. 
In various nematocerous groups the parascutella is more or less 
extended ventrad, ending near the base of the haltere or the 
metaspiracle. Among the Mycetophilidae this sclerite is more 
extended ventrad and beyond the metaspiracle. Especially in 
such mycetophilid gnats as the species of Mycetophila, Diomo- 
nus, Platyura, Asyndulum, and certain other genera, this sclerite 
deeply intrudes ventrad or ventrocephalad into the mesoepime- 
ron, and sometimes is almost contiguous with the base of the 
mesocoxa. As a result of the extreme progression of this ten- 
dency, the following unique sclerotization of the present insect 
is considered mainly to have been derived : The caudal half of 
the pleural side is almost entirely represented by the broad 
expansion of the parascutella; the ventral region of the post- 
scutellum is unusually expanded ventrad, forming a broad thick- 
ened wall of the sternal side and directly contiguous with the 
parasternite; the mesonotepimeron is nearly pushed out of its 
normal position by the cephalic extention of the parascutella; 
and the mesosternepimeron is widely separated from the meso- 
notepimeron by the deep cephalic intrusion of the parascutella 
along the epimeral suture. Another peculiarity derived from 
this unique parascutella is the relative position of the posterior 

• A similar feature is shown in certain Hymenoptera, Pepsia sp., by Snod- 
grass (1909). 

" This phragma is probably derived not only from the mesopostphragma 
but also from the direct fusion of the narrow metaprephragma with the 
intruding mesopostscutellum, losing the intersegmental membrane. 

^®' 2 Tokunaga: A Nymphomyiid Fly 165 

two pairs of legs. In all the dipterous insects they are always 
located close to each other, while in Nymphomyia alone the 
articulations of the meso- and metacoxse are very widely sep- 
arated from each other by the broad extension of the ventral 
region of the parascutellae. 

Thus, in the Nymphomyiidse we can trace, though with some 
difficulty, the homologic relation or evolutional line of the 
postscutellum, but so far as the development of the parascutellse 
is concerned, the present insect is so different from all known 
recent Diptera that their affinity cannot be discussed. 

As is generally the case in insects, the Diptera, as a whole, 
are provided with a well-developed mesopleuron (including the 
prsecoxale), which is clearly divided by the pleural suture into 
the episternum and epimeron. The episternum is usually sub- 
divided into the notepisternum and the sternepisternum by the 
episternal suture. The epimeron is also subdivided into the 
notepimeron and sternepimeron by the epimeral suture. In the 
present insect the pleuron is vestigial, and its major portion is 
represented by a uniform membrane, so that these sutures are 
almost invisible, the pleural suture (Plate 2, figs. 5 and 7, P2) 
alone being faintly represented by an incomplete, slightly un- 
dulated, creaselike line. In this vestigial and membranous con- 
dition the thorax of the present insect differs greatly from that 
of the Diptera in general, but is rather similar in appearance 
to that of the Apterygota or the immature forms of other in- 
sects. The pleural suture is always distinctly present on the 
dipterous thorax and serves as an important index in the phylo- 
genetic consideration of the Orthorrhapha as pointed out by 
Crampton. The nematocerous groups alone have more or less 
straight pleural sutures, while the brachycerous and atherice- 
rous insects have always sharply angulated sutures. The pleu- 
ral suture of the present insect simulates the condition of the 
former type, and this fact, even by itself, entitles this insect 
to a place in the nematocerous group, as I have already sug- 
gested in connection with other structures. 

The notepisternum becomes more or less membranous in va- 
rious nematocerous insects and sometimes it is completely re- 
duced into a uniform membrane, which is continuous with the 
rotaxis as seen in the thorax of D enter ophlebia and Edwardsina, 
but this tendency taken alone has no particular phylogenetic 
significance in the nematocerous group, since such a reduction 
of the notepisternum occurs sporadically in varying degree 

166 The Philippine Journal of Science 1935 

within the same major group or in closely related groups. 
Contrary to the above feature of the notepisternum, the steme- 
pisternum is always highly developed in the dipterous thorax, 
and the wide distance between the anterior two pairs of legs 
ia chiefly due to the development of this sclerite. Moreover, 
among the nematocerous groups the sternal area between the 
fore and middle legs is often completely or almost entirely 
represented by the ventrally expanded region of the stemepister- 
num (formed by the fusion of the episternum proper and the 
prsecoxale) in place of the mesosternum as seen in Tanytarszts, 
Chironorwm, Microtendipes, and other midges. In the present 
insect not only the notepisternum but also the stemepisternum 
has become membranous, and only the ventral margin of the 
pleuron is chitinized. Furthermore, the stemepisternum (Plate 
2, fig. 7, 6*2) is not extended ventromesad, differing in this 
respect from that of the generalized dipterous thorax, but re- 
mains in the proper position as in certain other Panorpoidea 
(Trichoptera and Lepidoptera) , being widely separated from 
that of the opposite side on the ventromeson by the well-developed 
mesosternum and its accessory membranous area. 

With regard to the epimeron of the present insect (Plate 
2, fig. 5), as already suggested, the notepimeron (712) and steme- 
pimeron (5^2) are very small and completely separated from 
each other along the epimeral suture by the deep cephalic in- 
trusion of the parascutella, which reaches the pleural suture. 
This complete separation of the epimeron (em2) is also a peculiar 
character for a dipterous insect since the two subdivisions of 
the epimeron, however widely they may be separated by the 
intruding parascutella, are always directly contiguous with 
each other along the cephalic corner of the parascutella, as in 
the higher Nematocera (Mycetophilidae and Bibionidae) and the 
lower Brachycera (Leptidae). The notepimeron of the present 
insect contracts cephalad and becomes a small sclerite, which 
appears superficially to be one of the basal wing sclerites. This 
condition is brought about by the cephalic intrusion of the large 
parascutella, which has become directly contiguous with the 
cephaloventral margin of the large pleurotaxis and shifted the 
postalifera caudad along this margin. 

Thus, the great modification and peculiarity of the meso- 
pleural sclerotization of this insect are chiefly due to the ex- 
tremely anomalous development of the parascutellse accompanied 
by the membranation of the pleuron itself. 

^' ^ Tokunaga: A Nymphomyiid Fly 167 

According to Snodgrass the typical definitive sternum (text 
fig. 3) consists of the three primary sclerites, presternum (Ps), 
basisternum (Bs), and sternellum (SI), and two kinds of the 
secondary sclerites, spinasternum (Ss) and paired laterosternites 
(Ls). These sclerites are often highly variable in form, posi- 
tion, and extent of the development among closely allied groups, 
and study of their homology is extremely difficult, especially in 
the highly modified thorax such as of the Diptera. On the 
mesothorax of the lower nematocerous Diptera, Tipulidse, 
Crampton has pointed out the following sternal sclerites: The 
presternum, basisternum, prefurcasternum, postfurcasternum, 
and laterosternites. The terminology of these writers differs, 
but it seems that the presternum of Snodgrass includes both the 
presternum and the basisternum of Crampton, that the basi- 
sternum and the sternellum of the former are homologous with 
the prefurcasternum and the postfurcasternum of the latter, 
respectively, and that the spinasternum is quite original with 

In the case of the Nymphomyiidse the mesosternum consists 
of two distinct sclerites. The cephalic sclerite (Plate 2, fig. 
7, prs2), the so-called sternannum,ii is conspicuously separated 
from the caudal one (Plate 2, fig. 7, bs and snh), the so-called 
sternellum, by the membranous transversal area (ss), which 
is provided with a pair of small setae (these setae are very dif- 
ficult to detect on account of their delicate structure). The 
setigerous nature of this membranous area suggests that it is 
different in origin from the furcinae or their derivatives and 
probably a secondary development, where flexibility is demanded. 
If so, these sclerites cannot represent the sternannum and ster- 
nellum of MacGillivray. The presternum (Snodgrass, text fig. 
3, Ps), which is clearly shown in the thorax of the Embiidae 
(CylindrachMa) , is more probably homologous with this cephalic 
sclerite (Plate 2, fig. 7, prss) of the present insect. In the Ti- 
pulidse, the sclerite homologous with the presternum (Snodgrass) 
consists of the presternum and basisternum (Crampton), but 
in the present insect such secondary subdivision is found neither 
externally nor internally on the cephalic sclerite, and it is al- 
ways represented by one large uniform chitinization. The 
caudal sclerite of the present insect is provided with a pair 

^MacGillivray divides the sternum into two sclerites, such as the ster- 
nannum and sternellum, according to the position of the furcinae. 

168 The Philippine Journal of Science i938 

of small sternacoilse (SC2) and a V-shaped invagination of the 
furcag on its cephalic region. These structures of the caudal 
sclerite show that it is homologous with the furcasternum 
(Crampton) . The furcasternum or its homologous sclerite, the 
primary sternum, is subdivided by a thickened transversal in- 
vagination (text fig. 3, fs) into the pre- and postfurcasternum 
(Crampton) or the basisternum and sternellum (Snodgrass). 
This subdivision can be followed in the small V-shaped mem- 
branous area and the caudal elongated thickened area of the 
present thorax. Concerning the origin of the longitudinal ental 
ridge (usually known as the midventral suture) of the caudal 
sclerite, two hypotheses may be advanced. This ridge may 
have been derived from the median spina of the spinasternum 
through its cephalic extension, or its origin may be traced back 
to the common invagination (the V-shaped transversal thicken- 
ing or transternal suture) of the furcae, which has become ex- 
tended caudad accompanied by the ental invagination of the 
sternellum, as in the "Sternagrat" (Weber) of the basisternum 
of many pterygote insects. The former hypothesis considers 
the caudal sclerite as a fusion product of the primary sternum 
and the secondary spinasternum, but this secondary interseg- 
mental sclerite occurs only in the thorax of certain nymphs 
and of certain lower Orthoptera and is completely wanting in 
Diptera in general. This longitudinal invagination of the Dip- 
tera extends throughout both the meso- and metafurcasternum 
and is always completely continuous with the common invagina- 
tion of the furcae, both externally or in the sutural structure, 
and internally or in the apodemal structure. Thus, in so far 
as this longitudinal ridge is concerned, the latter hypothesis, 
in which the primary sternum is secondarily modified by the 
development of an apodemal brace derived from the furcae 
where rigidity is demanded, seems more natural, and therefore 
more reasonable than the former hypothesis. 

The laterosternites (s. str.), which originate from the sub- 
coxal elements, are found around the coxafossae in the hypo- 
thetical thoracic segment, but in almost all insects, except the 
Apterygota and nymphal forms, and especially the Diptera, 
they are firmly fused with the primary sternum, even in the 
Protonematocera. In the present case the laterosternites are 
also regarded as completely fusing with the caudal elongated 
sclerite instead of being completely atrophied, since the trace of 
the subcoxal elements (special parts of the laterosternites) is 

^*' ^ Tokunaga: A Nymphomyiid Fly 169 

shown as a pair of slender lateral thickenings, the precoxal 
bridges or pr^coxales, which are located cephalad of the coxa- 
fossse and united with the primary sternum proper. 

To summarize, this sclerite is considered as consisting of the 
two divisions of the definitive sternum, namely, the basisternum 
and sternellum, and the subcoxal elements, namely, the latero- 
sternites and precoxal bridges (text fig. 3) . 

Besides the sternal sclerites already discussed, there is, on 
the sternal membrane, closely along the ventral margins of the 
mesoparascutellse, a pair of thickened areas that in the present 
paper are conveniently described as the parasternites (Plate 
2, fig. 7, pt). The homology and origin of these sclerites is 
difficult to discern. However, in all probability they were de- 
rived either from the secondary chitinization of the sternal 
membrane or from certain postscutellar sclerites homologous 
with the katapleurotergites (Young) .^^ 

The external chitinization of the metanotum is completely 
wanting or greatly reduced secondarily in the Diptera, although 
rarely distinct and less modified as in Psychoda and the allied 
genera, and generally it is represented by a very narrow semi- 
chitinized band of integument connecting the bases of the hal- 

" The parascutella (MacGiUivray) or pleurotergite (Crampton) is subdi- 
vided by Young (1921) into two sclerites: the anapleurotergite and kata- 
pleurotergite. Although Young did not discuss the distribution of these 
sclerites among the Diptera, they are shown by his figures in the following 
groups of the Orthorrhapha : the Calypterae (excepting the Dexiidae-r/ie- 
laria) and the Acalypterae (excepting the Micropezidae-Caiobata, Piophi- 
lidae-Piophila, Ephydridsd-Parydra, and Oscinidse-C/i^orops) and never 
shown in the lower Diptera such as the Athericera, Brachycera, and Ne- 

Each lateral portion of the mesascutella is subdivided by MacGillivray 
(1923) into two sclerites: the parascutella and durascutella. He said 
that the parascutella sometimes extends onto the pleuron and forms a 
distinct area, as in the Diptera, between the mesepimeron and the mete- 
pimeron, and that the durascutella is the infolded longitudinal thickening 
on each side separating the mesascutella from a parascutella. From his 
statement the parascutella may be homologous with the pleurotergite 
(Crampton) or the ana- and katapleurotergites (Young) taken together. 
Hence the ana- and katapleurotergites cannot be homologous with the 
dura- and parascutellaa, respectively. The durascutella of the dipterous 
thorax has been long neglected by many writers, but I reported the pres- 
ence of the homologous sclerite on the postscutellum of a certain marine 
crane fly, Limonia sp., in 1930, and I found thereafter wide distribution of 
this sclerite among various crane flies where it is developed in different 


The Philippine Journal of Science 


teres or by an intersegmental membrane between the infolded 
mesopostscutellum and the precostal thickening of the first 
abdominal segment. The metanotum of the present insect dis- 
tinctly belongs to the latter case, losing completely both the 
external and the internal chitinization, due to the secondary 
modification (Plate 2, fig, 11). 

The metapleuron (Plate 2, fig. 5, eta and ems) is directly 
adjacent cephalad to the mesoparascutella (PS2) in the present 
thorax, due to the extraordinary development of the latter, in- 
stead of being adjacent to the mesopleuron as is usual in insects, 

Fig. 8. Hypothetical definitive sternum. (Modified from Snodgrass, 1981.) acx, Precoxal 
bridge; Bs, basistemumj ex, coxa; /«, f ureal suture; Ls, later osternite ; pcoj, postcoxal 
bridge; P», presternum; sa, sternartis; «c, sternacoila; SI, sternellum; sptt, spinafurca; 
9ps, secondary presternal suture; Sa, spinastemum. 

although the position of the metapleuron remains normal. The 
relative position between the episternum and epimeron of this 
thoracic segment is almost normal. However, the episternum 
(etg) is completely membranous, hence the metapleuron appears 
to be represented by the undivided pleural sclerite. The epi- 
meron (ems) is somewhat larger than the membranous epister- 
num. This tendency, although distinctly shown in a rare case 
(Dolichopus) , is not usual among the Diptera, where the de- 
velopment of these sclerites in size is reversed. 

The metasternum (Plate 2, fig. 7, sm^) is represented by a 
large sclerite that is homologous with the caudal elongated 
sclerite of the mesothorax. Although the shape of this sclerite 

^^' ^ Tokunaga: A Nymphomyiid Fly m 

is different, the structures are very similar to those of the 
mesothorax. It differs only in the presence of the postcoxal 
bridges, which are completely wanting in the mesothorax. The 
sclerite homologous with the large presternum and the small 
basisternum of the mesothorax is completely reduced to a uni- 
form membrane in the metathorax, hence a large membranous 
area persists between the two caudal chitinized sternal sclerites. 
As stated above, the chitinized metasternite, which is the only 
chitinized part of the metasternum, is a caudal half of the 
definitive sternum, and fundamentally different in the mor- 
phological significance from the undivided prothoracic sternum 
ivst), since the latter represents the whole of the definitive 

The following is a summary of the principal conclusions 
derived from the study of the thoracic sclerotization of the 
Nymphomyiidae : (a) The peculiar general appearance of the 
thorax is chiefly due to the short but broad membranous condi- 
tion of the cervix, the extreme elongation of the scutoprsescutum 
and the mesascutella, the loose arrangement of the large sternal 
sclerites, the rudimental condition of the pleuron and the ster- 
num, and the extreme development of the parascutellae. (6) 
The sclerotization of the notum coincides well with that of the 
general dipterous thorax, (c) The wide space between the 
middle and hind legs and the displacement of the mesopleural 
sclerites are both probably connected directly with the anoma- 
lous expansion of the parascutellae. (d) The extreme develop- 
ment of the sternal sclerites is thought to be chiefly due to the 
rudimental condition of the pleura, and has resulted in the ano- 
malous lateral position of the forelegs, {e) The loose arrange- 
ment of the sternal sclerites is considered to be due partially to 
the secondary reduction of the preexisting sclerites and partially 
to the rudimentary condition of the adjacent sclerites. (/) 
Although the sternal sclerotization appears to be very primitive, 
it is not sufficiently primitive to be compared with that of the 
nymphal forms or lower orders. It probably represents the 
most primitive condition among the known dipterous insects. 
{g) The nonangulated pleural suture supports the assumption 
that this insect belongs to the nematocerous Diptera. From the 
primitive character of the sternal and pleural sclerites this 
family seems to be placed in a comparatively low phylogenetic 
position in the Nematocera, but the characteristic modification 
of the various sclerites suggests that this insect may be far 

172 The Philippine foumal of Science 1935 

more advanced than the hypt hetical ancestors of Diptera, and 
represents a specialized fami^ isolated from all other known 
dipterous groups. 

Although in the previous report (1932) the thoracic spiracles 
were stated to be absent, further study has brought to light 
very minute, porelike, paired punctures (Plate 4, figs. 11 and 
13, sp), about 2.5 iJt. in diameter, situated at the position where 
the spiracles would be expected. These punctures are very 
much smaller than the true spiracles of the Psychodidse and 
allied insects, which are known to have extremely small thoracic 
spiracles for the dipterous insects. The external structures, 
such as the peritreme and related thickenings normally asso- 
ciated with the spiracles, have not been detected in the study 
of the punctures under consideration, and it is highly question- 
able whether they are functionally open. 


The shape of the wings is very diverse among the Diptera. 
Elongation and narrowing frequently occur synchronously with 
the elongation and narrowing of the body and various append- 
ages, and this relationship is most commonly found in the fol- 
lowing Nematocera: The Tipulid^, Dixidse, Culicidae, and Chi- 
ronomidse. The higher groups are usually provided with broad 
wings proportional to the massive body but rarely elongated. 
Narrow wings are shown in the following slender flies: The 
Leptogastrinse and Dioctria (Asilidse), Systropinse (Bomby- 
liidae), Xylophaginae (Leptidae), Nerius and Calobata (Micro- 
pezidse), and Baccha (Syrphidae). These features of the wings 
are derived from the reduction of the caudal area, squama, and 
alulae, if present, of the wing membrane. As the result of these 
reductions the caudoproximal margin of the wing^^ becomes 
more or less straight and nearly parallel to the anal or cubital 
vein, as shown in the wings of certain Systropus (Bombyliidse) , 
Euglochina (Tipulidse), and StempelUna (Chironomidae). Also 
in the Nymphomyiidse the elongation and the narrowing of the 
wings well coincide with the above process of the reduction. 
However, the following features of the wing shape seem to be 

"In certain Hymenoptera, Homoptera, and Lepidoptera the forewings 
become very narrow and straight at the caudal margin, losing the large 
anal lobes. This is due to the development of a special structure that fas- 
tens together the two wings on one side for the synchronous wing action. 
Thus the reduction and narrowing of the forewings of these insects are 
somewhat different in significance from those of the dipterous wings. 

»8' 2 Tokunaga: A Nyy^^phomyiid Fly 173 

quite characteristic for the presef f: insect; namely, the extreme 
reduction of the squamal area ifad the sharply angulated or 
pointed wing tip, since in those dipterous insects provided with 
elongated wing the distal area, distad of the cord veins, is usually 
blunt or round, and the squama remains normal for the family 
or genus to which they belong. 

The long setal fringe of the wings is another characteristic 
structure of the present insect. The wings of the insects, as a 
rule, are more or less fringed with hairs, but they are usually 
inconspicuous and very short in proportion to the size of the 
wings. The long hairy or scaly fringe is never found irregularly 
or at random, but its occurrence is limited to the following 
groups : The order Thysanoptera ; certain parasitic Hymenoptera, 
the Trichogrammidse and Mymaridse; small Coleoptera, the 
Ptiliidae; and tineinid Lepidoptera, the Gelechiidse, Cosmoptery- 
gidae, Elachistidse, Heliodinidse, Gracilaridae, Lyonetiidse, Coleo- 
phoridse, etc. Among the dipterous groups the fringe, although 
far shorter than in the above insects, is found in the Culicidae, 
Psychodidae, Ceratopogonidae (Culicoides, Forcipomyia, Atricho- 
pogon, Lasiohelea, and Dasyhelea), Cecidomyiidae (Bryocrypta, 
DicroneuvTis, Holoneurus, Diallactes, Winnertzia, and Monodi- 
ptosis), and Chironomidae (Zavrelia, Lauterborniella, Polype- 
dilum, Tanytarsus, Stempellina, Metriocnemus, and Cardiocla- 
dius). Of the above dipterous groups the development of the 
wing fringe of the first three families is closely associated with 
the development of the common setae or scales of the wing mem- 
brane, and no special differentiation, in size or structure, is 
shown between the setae on the fringe and those on the membrane. 
In the last two families, although the fringes are not so con- 
spicuous when compared with the extreme development of the • 
marginal setae of the present insect, they show more or less 
special differentiation as an independent structure, the common 
hairs (both the micro- and macrotrichia) of the membrane 
being completely or highly reduced, both in structure and in 

In general the long setal fringes are well developed in the 
wings of minute insects, as mentioned above; the length is 
usually about or less than 2 to 3 millimeters in dipterous flies ; 
2 to 3 millimeters in thrips, except certain tubuliferid thrips, 
which rarely reach 7 millimeters in body length; less than 2 
millimeters in chalcid flies, which include probably the smallest 
hymenopterous insects (0.21 millimeter in body length) ; and 
ptiliid or trichopterygid beetles, which contain the smallest known 

174 The Philippine Journal of Science isse 

beetles, ranging down to 0.2 millimeter. Among the Lepidop- 
tera the long fringes are found in forms larger than in other 
orders, about 8 millimeters or less in body length, but as their 
designation, Microlepidoptera, signifies, these moths are small 
compared with lepidopterous insects in general. Furthermore, 
the setal fringes are the better developed the narrower the 
wings ; they are best exhibited in small tineinid moths. For ex- 
ample, among the Yponomeutidse, Plutellidse, Glyphipterygidse, 
and Tischeriidse, species of the narrower-winged genera Argy- 
restia, Plutella, Lamprystica, Glyphipteryx, and Tischeria are 
usually provided with longer fringes than species of the broader- 
winged genera, Yponomeuta, Ethmia, Cerostoma, Niphonympha, 
Imma, Tortyra, Simaethis, Brenthia, Choreutis, and Solenobia^ 
In the fissured wings, like those of the Pterophoridae and Omeo- 
didse, the marginal fringes are also highly developed, but in the 
same family the genus Ochyrotica, which has nonfissured massive 
wings, has shorter wing fringes than the other fissured-winged 
genera. Among the Diptera this tendency, although not so 
conspicuous as among the lepidopterous insects, is also exhibited 
among the Chironomidae ; the wing areas in the genus Tany- 
tarsus are reduced in various degrees in different subgenera; 
for example, species in Stempellina and Zavrella, which are 
provided with rather narrow reduced wings that have better- 
developed marginal setae than those in other subgenera, such 
as Micropsectra, Tanytarsus, and Lundstromia, which are pro- 
vided with rather broad wings. Moreover, this tendency is 
shown even among species of the same subgenus ; for example, 
in the subgenera Stempellina and Sndtta, St cuneipennis Ed- 
wards, St. hrevis Edwards, Sm. angusta Edwards, etc., which 
.have the characteristic narrow wings, are provided with longer 
setal fringes than the broad wings of other species of the re- 
spective subgenera. 

The facts that the long setal fringes are well developed in 
minute narrow-winged insects, ''microes,*' which are feeble 
flyers, and that they are never found among large active broad- 
winged insects, which are swift flyers, suggest that the function 
of these long fringes is to increase the buoyancy of the insect 
in the air; that is, to increase the soaring coefficient, as suggested 
and demonstrated by V. N. Chitrovo (1914) and C. W. Collins 
(1915) in their studies of plant seeds and the gipsy-moth larvae, 
respectively, rather than to increase the efficiency of the wings 
as propelling organs by compensating for the reduced wing area. 

^^ Tokunaga: A Nymphomyiid Fly 175 

The wing venation of this insect reveals a condition of high 
specialization, which is in accord with the degree of develop- 
ment of other structural features. Among the Diptera the vena- 
tion of the Tipuloidea and Psychodoidea represents a primitive 
stage, while that of the Chironomidse, Ceratopogonidse, Simu- 
liid«, Cyrtidae, Phoridae, Hippoboscidse, etc., is more specialized, 
in that the cephalization and the thickening of the veins 
have taken place in varying degrees. The latter group of dip- 
terous insects is generally recognized, as a whole, as more 
specialized than the protonematocerous group, which is provided 
with the tipuloid or psychodoid venation. Viewed from this 
standpoint, the proximocephalic concentration of cephalic veins 
and the reduction of caudal veins of this insect suggest that the 
species under consideration is fairly high in the evolutional 
scale and quite advanced in its own direction of specialization. 

The presence of the well-developed ambient vein (veinlike 
structure of the wing margin) in the wing of this insect is one 
of the peculiar features common to the Nematocera in general. 
The ambient vein is considered homologous with the ''ring vein" 
of the Thysanoptera and develops quite often in connection with 
the wing fringe, compensating possibly for the reduction of wing 

The entire proximal group of veins of the present wing ends 
on the costal margin. Among the dipterous insects the veins 
that end on the costal margin are either the subcostal or the 
radial veins or both. The subcosta often becomes very short 
or even completely atrophied, and more rarely coalesces with the 
radial vein ; if present, it is generally retained as an independent 
vein. In the present wing the double nature at the proximal 
part of the costal margin shows that this portion consists of 
the costa coalesced with the subcosta. This feature of the co- 
alescence is peculiar, not only in the Diptera but also in the other 
orders, since the subcosta, as mentioned above, usually coalesces 
with the radial vein instead of the costa, as shown in the fore- 
wings of certain Hymenoptera and Homoptera, and in the hind 
wings of the heteroneurous Lepidoptera in general and certain 
Neuroptera. The similar or related features are occasionally 
found in other orders, such as the Jassidae (Gypona) and Psyl- 
lidae (Psylla), and also in certain ichneumonoid Hymenoptera, 
the Ichneumonidae and the Braconidae, where the condition 
of C + Sc + R obtains. 

176 The Philippine Journal of Science i936 

The radial vein is the most stable wing vein and is distinct 
even in such highly reduced venation as that of the Embioptera 
(Oligotoma) , Thysanoptera, Homoptera (both sexes of the Aleu- 
rodidsB and males of the Coccidse), and Diptera (Cecidomyiidae, 
Simuliidae, and Phorid^) . Moreover, the radius of the general- 
ized wing is provided with a definite branch, the radial sector; 
among the Diptera this sector, although subdivided into from 
two to four branches, is always more or less angulately branched 
off from the straight stem from which Ri extends straight for- 
ward. On the basis of these general characters of the radial 
vein — ^the characteristic position, its constancy, and the feature 
of branching — ^the second cephalic group of veins of the present 
insect are likely to be homologous with the radial vein and its 

Of this radial group of the present insect the cephalic straight 
vein is evidently homologous with the first branch of the radius, 
but it is questionable whether the caudal branch is homologous 
with the radial sector itself or With vein R44-5. The two- 
branched condition of the radial sector is universally found 
throughout both the orthorrhaphous and cyclorrhaphous groups 
of the Diptera. These two branches, R2+3 and R4+5, are de- 
finitely present, however reduced the other veins may be, among 
members of the higher dipterous groups, but very rarely among 
the Nematocera R2-1-8 degenerates in situ into a very feeble 
obscure vein, although not completely obsolete, as shown in the 
ChironomidsB. However, the simple condition of the radial sec- 
tor is frequently found among the Orthorrhapha, such as the 
CeratopogonidsB, Simuliidae (excepted Prosimulium) , Cecido- 
myiidse, Mycetophilidse (Sciarinse and Mycetophilinse) , and Cyr- 
tidse. This simple condition is derived from the unbranched 
Rs or from the distal coalescence of veins R24.8 and R4+5. There- 
fore, the cephalic and caudal veins of the radial group of the 
N3nnphomyiid3e should be designated as Ri and Rs, respectively. 
In my previous taxonomic report the latter vein was mentioned 
as R44.5, R2+8 having been considered to be completely atrophied. 

The most caudal vein of the present fly may be homologous 
with either the cubital or the anal vein when considered in 
connection with its position. However, when the reduction 
occurs in the anal area of the wing, the anal veins atrophy at 
first and the cubital veins remain more or less stable, as shown 
in the wings of certain Termitidse, Oligotomidse, Zorotypidae, 
Thysanoptera, Aphididse, and Aleurodidae. The opposite case is 


Tokunaga: A Nymphomyiid Fly 


not definitely known. This tendency is also found in the dip- 
terous groups that have the reduced anal lobes, like certain 
Systropinae, or have reduced venation, like the Mycetophilidse, 
Cecidomyiidse, and Cyrtidse. Moreover, the comparatively broad 
nature of the present vein, although it is not sharp and strong, 
is another important character for the suggestion of homology 
with the cubital vein, since the two cubital veins, Cui and Cug, 
always lie in a close parallel position, apparently forming one 
broad strong vein in the general dipterous wings. In this sense 
the most caudal vein of the Nymphomyiidse may have been 
derived mainly from the coalescence of the two cubital veins, 
Cui and Cus. 

The median veins of wings are the most unstable veins, being 
frequently reduced or degenerated highly or completely, as we 
see in the wings of the Embioptera, Thysanoptera, and certain 
oligoneurous Homoptera and Diptera. This feebleness and the 
characteristic position between the radius and cubitus suggest 


Fig. 4. Hypothetical venation. Sc, Subcosta ; Ri, first branch of radius ; Rs, radial sector ; 

M, media ; Cu, cubitus. 

that the creaselike vein of the present insect may be homologous 
with the vein media, unless it is derived from the folds 
as in the Blepharoceridae and Deuterophlebiidse. Whether this 
vein represents the unbranched media itself or either vein M^^a 
or M3f4, is highly problematic, since both conditions are found 
even in the same dipterous group. 

The hypothetical wing venation of the present fly is shown in 
text fig. 4. It distinctly shows the nematocerous characters in 
the following points; namely, the median cell and the closed cell 
Cui are wanting. Among the nematocerous wings there are 
two types of venation, the protonematocerous and eunematoce- 
rous types. In the former type the differentiation of the veins 
is very poor, and the radius and media are provided with many 
branches, as in the wings of the Psychodoidea (Tanyderidae, 
Ptychopteridse, and Psychodidse) and Tipuloidea (Trichoceridae 

289853 6 

178 The Philippine Journal of Science im 

and Tipulidse), while in the latter type the wing venation is 
more differentiated and the radial and median veins are provided 
with only a few branches, as shown in the Chironomoidea, Bi- 
bionoidea, Mycetophiloidea, etc In this respect the present 
venation appears more closely similar to that of the Eunemato- 
cera than to that of the Protonematocera. Certain genera of 
the Cecidomyiidae, Mycetophilidae, and Chironomidse show a 
reduction of the wing veins similar to that of the Nymphomyiidse 
in the simple condition of the radial sector, the median, and the 
cubital vein.^* 

The wing base is very difficult to study, because the small 
sclerites are easily shifted from their normal positions and as- 
sume very different appearances. This is especially true in the 
case of minute flies, such as the present species. On the panor- 
poid wings the basal sclerites are generally well developed and 
their typical arrangement is well shown on the mecopterous 
wings. In the Diptera their arrangement and development are 
more or less modified, but those of the Tipuloidea are almost 
typical, as shown by Snodgrass (1909), Crampton (1919), and 
myself (1930). These sclerites of the present insect are greatly 
reduced, both in size and in number, being correlated with its 
feeble flying habit. Their arrangement and structure, however, 
suggest that they still allow the dorsoventral fluttering motion 
of the wing. 

Of these basal sclerites the tegula is usually highly developed 
in the Diptera, Hymenoptera, Lepidoptera, and Mecoptera, and 
this sclerite is shown even on the base of the dipterous haltere. 
In the case of the Nymphomyiidse it is vestigial and far removed 
from the tip of the costalis (Plate 4, fig. 13, f 62) . 

Of the five aliferae the pleuralifera or the pleural wing pro- 
cess always persists on the thorax of the Pterygota and consists 
of a prominent pillarlike projection of the pleuron, which is 
divided into two parts by the pleural suture. But in the dip- 

" The cubital forks, which are formed by the anastomosis of two adjacent 
veins, Mh-4 and Cui, usually persist on the wings of the Eunematocera ; how- 
ever, they are frequently reduced into a simple condition of Cui as in the 
present venation, either by the complete reduction of the cephalic branch, 
Ms-H, as in certain MycetophilidsB (Monoclona, Acnemia, Azana, Zygom- 
yia, and Sceptonia) and Cecidomyiidae {Holoneura, MiasJter, Trisopsis and 
Diallactes) or by the incomplete anastomosis or by the reduction of the 
crossvein, m-cu, as in the general Simuliidae, certain Mycetophilidae 
(E'Ctrepesthoneura, Plastasciara and Sciara) and Cecidomyidae {Winnert- 
zia and Clinorhytis) , 

5«» 2 Tokunaga: A Nymphomyiid Fly 179 

terous pterothorax the above original condition of this alifera 
is more or less modified by a partial reduction of the episternal 
or epimeral sclerites as well as by the additional development 
of the secondary sclerite, the subalifera (Crampton). In the 
case of the simplest modification the pleuralif era is represented 
only by either the episternal region, losing the caudal half, as in 
the Dixidae and Tanyderidse ; or the epimeral region, losing the 
cephalic half, as in the Blepharoceridse and Limnobiinse (Tipu- 
lidse). A higher modification, which is very common among 
the Nematocera, occurs by the addition of the subalifera in 
the position of the episternal region of the pleuralifera. In 
the highest modification the pleuralifera is entirely represented 
by the subalifera itself, completely losing the epimeral region. 
In this state the pleuralifera is completely separated from the 
pleuron and exhibits an independent islandlike structure similar 
to the other aliferse freely located on the pleurotaxis, as shown 
in the Anisopodidse, Leptidae, Thaumaleidse, etc. From the 
above morphological point of view the independent pleuralifera 
(Plate 4, fig. 13, pr) of the present insect must no doubt be 
included in the last category. 

Of the other aliferse the most caudal micralifera is always 
wanting or fused with the postalifera in the higher Pterygota, 
and it is also wanting in the present insect. The three inde- 
pendent aliferae (the pre-, med-, and postalifera) are always 
present in the Pterygota, and the largest of the three is provided 
with a large ental disc, which serves as an attachment of the 
wing muscles. This largest alifera is not limited to the specific 
sclerite, but is represented by either of the above three aliferae. 
This ental disc, the pronator disc (Amans), is usually present 
on the medalif era in the Coleoptera, Lepidoptera, and Hymenop- 
tera, while in the Diptera it is largest on the postalifera, as I 
have already pointed out in the study of a certain crane fly 
(1930) . Although the medalifera (Plate 4, fig. 13, ml) is highly 
modified into a thin membrane in the present insect, these three 
aliferae (pal, ml, and pol) can be distinguished as in the gen- 
eralized Diptera, but the pronator disc is not developed on any 
of them. This simple condition of the postalifera also suggests 
that this fly is not an active flyer. 

The pteraliae are classified into the following groups by Mac- 
Gillivray (1923) : The duritse, funditse, and venellae. The du- 
ritse, which are also known as the axillaries, hypothetically 
consist of four sclerites, the sigmoidea, submedia, terminalia. 

180 The Philippine Journal of Science 1935 

and navicula. In the pterothorax of the Diptera the first scle- 
rite is most strongly developed of the four, being provided with 
an ental disc to which the wing muscles are attached, and pos- 
sessing a neck region against which the costalis is directly 
articulated; while in the present insect these duritae are ob- 
solete, being represented only by the terminalia (Plate 4, fig. 13, 
t). The costalis is here loosely articulated to the notal wing 
process (caudalaria?) accompanying a minute obscure scle- 
rite, which may be either the reduced sigmoidea or submedia. 

The funditae, or median plates (Snodgrass), are very unstable 
sclerites, being easily reduced or united with the venellse. In 
certain orders, Orthoptera, Coleoptera, Hemiptera, etc., these 
sclerites are always present, while in the Lepidoptera and Hy- 
menoptera they are generally obscure or wanting, and when 
present they are represented by only one sclerite, the mediella, 
as in the lower Nematocera (Limonia) . It must be noted here 
that the term "median ossicles or medipterales" used by Cramp- 
ton (1919) in the study on the wing-base sclerite of the Pan- 
neuroptera (Hymenoptera, Neuroptera, Mecoptera, Diptera, 
Trichoptera, and Lepidoptera) is somewhat different in meaning 
from the term "funditae or median plates," since his "median 
ossicles" include the various duritse other than the sigmoidea 
in addition to the true funditse. 

In the present wing three venellse (Plate 4, fig. 13), costalis 
(cs), radialis (ra), and analis (as), are distinctly present as 
in the generalized wing base, and these structures are highly 
thickened as compared with the other reduced pteraliae. Their 
arrangement and relationship to the wing veins have provided 
important evidence for the homology of the wing venation. 

It is very peculiar that the wings of this fly are deciduous, 
being easily broken off along a definite line, as is commonly 
known in the Isoptera and Formicidse. Whether or not this is 
a normal occurrence in life is still problematic. The wing is 
especially thinned between the highly thickened venellse and the 
wing stem, showing a suturelike line. When the wings break 
off along this line, the wing stumps resemble those of ants rather 
than the scapular shields of termites. Deciduous wings are rare 
among the Diptera, being known only in Nymphomyia and in 
the female of the parasitic fly Ascodipteron (Streblidae), in 
which the remaining portion of the wing also consists mainly 
of the venellse, as in Nymphomyia. 

5«- 2 Tokunaga: A Nymphomyiid Fly 181 


Delicate and slender legs like those in the present insect are 
prevalent in the nematocerous group, but the relative length of 
the three pairs of this fly is not common in dipterous insects. 
Among the Diptera, as a rule, the hind legs are the longest 
and the middle shortest. Rarely the forelegs are the longest 
and the middle shortest, as in the Chironomidse. In the present 
fly the forelegs are the longest and the hind shortest. More- 
over, the legs usually become elongated, paralleling the slender- 
ness of the abdomen, wings, and various external appendages, 
as is best exhibited in various families of the Nematocera. In 
the Nymphomyiidse the legs are very small for the size of the 
insect or foi" its extremely elongated abdomen. 

The elongated coxae and trochanters of Nymphomyia are also 
unusual characters for the Diptera and rarely found in the 
nematocerous group. The elongation of the coxae is most dis- 
tinctly shown in the archaic psychodoid forms, such as Brur- 
chomyia and Nemopalpus, which probably belong to the oldest 
existing types of Diptera. The Mycetophilidae (excepting the 
Sciarinae) are also provided with elongated coxae. This family, 
although belonging to the eunematocerous group, retains many 
ancestral or primitive characters in certain structures of the 
antennae, the thorax, and, especially, the legs, which suggests 
that it is intermediate between the other panorpoid groups 
(Mecoptera, Trichoptera, and Lepidoptera) and the Diptera. 
The elongation of the coxae is also found in the forelegs of other 
archaic Nematocera, the Rhyphidae (Anisopodidae or Anisopidae) , 
which retain many structures related to the older fossil Mecop- 
tera, and it is assumed that the Eunematocera, Brachycera, and 
other higher Diptera have descended from a common rhyphidlike 
ancestor. The elongation of the coxae is usually found in archaic 
dipterous forms, which suggests the probability that they are 
more primitive phylogenetically than those that are shorter.^^ 

" The coxae often appear superficially to be distinctly elongated in the 
lateral aspect on account of the following modification: The lateral sides 
of the coxae become extended triangularly dorsad (Nematocera) or dorso- 
laterad (Brachycera and others), whereby the coxacoilse and artes are 
found at about the middle in height of the thorax, while the mesal sides 
remain in the original short condition. Thus this superficial elongation 
is due to the secondary deformation of the coxae and need not be dis- 
cussed here. 

182 Ths Philippine Journal of Science tns 

Although very rare in the Diptera, the secondary elongation 
of the coxae may occur along the line of specialization in con- 
nection with the high modification of the other structures of 
the legs. It is best exhibited in certain remarkable genera of 
the higher Nematocera, such as Canthyloscelis and Corynoscelis 
(both Scatopsidse) and in the highly specialized grasping legs 
(forelegs) of the Ephydridaa. 

Each coxa of this fly presents a well-marked basal rim set off 
by a narrow submarginal membranous ring. This basal rim is 
homologous with the basicoxite of the generalized legs of insects, 
and the coxal muscles directly start from this rim, not being 
provided with special apodemal discs (promotors and remotors) . 
The narrow membranous ring is homologous with the basi- 
costal suture (Snodgrass). In the lower primitive orders (Or- 
thopteroidea) this suture internally forms a strong basicosta, 
while in the higher orders it is very much obscured or entirely 
wanting. Among the Panorpoidea this suture is usually very 
much obscured or completely atrophied, although it is always 
present in Trichoptera, at least on the mesocoxa. In the Dip- 
tera this suture is almost always completely wanting. It is 
very strange that in spite of this the present fly is exceptionally 
provided with this suture on the coxa of each leg. 

The coxa of the pterothorax, as a rule, is subdivided into two 
parts, the eucoxa (veracoxa or coxa genuina) and meron. In 
the Panorpoidea generally this feature is distinctly shown in 
its primitive condition, but only the Diptera show a complex 
modification on this point. The subdivision of the dipterous 
coxa very rarely shows the primitive orthopterous type in which 
the coxal suture or alasuture longitudinally extends throughout 
the coxa from the trochacoila to the coxartis. This is exhibited 
by Pedicia (Young) and Limonia (Tokunaga), in the lowest 
family, the Tipulidae. Large members of the nematocerous 
group show the more or less modified coxal subdivision similar 
in appearance to that of the Mecoptera, where the coxal suture 
extends parabolically on one side of the coxa. Thus the prox- 
imal margin of the coxa consists of the two sclerites, the eucoxa 
and meron, but the distal margin is represented by the eucoxa 
alone. Further along this line of modification, the meron mi- 
grates dorsad, intruding onto the pleural region, and at last this 
sclerite becomes completely fused with the epimeron or sterne- 
pimeron, forming the meropleurite or merosternepimeron. Thus 
the first segment of the leg is represented solely by the eucoxa, 

^^'^ Tokunaga: A Nymphomyiid Fly 183 

exhibiting an appearance similar to the procoxa, which is not 
differentiated into the two sclerites. This condition is best ex- 
hibited among certain special Nematocera; such as, the Psycho- 
didae, Blepharoceridae, Deuterophlebiidae, Tanyderidffi, and Pty- 
chopteridas, and the other higher groups, such as the Brachy- 
cera i« and Cyclorrhapha. Often the meron is reduced to a uni- 
form membrane, as in certain chironomid, ceratopogonid, myceto- 
philid, and cecidomyiid flies. In this reduced state of the meron 
the coxa appears somewhat similar in structure to the primitive 
undivided condition as well as in the most highly modified form, 
since the coxa is represented by one chitinized sclerite, the eu- 
coxa, alone. Of these types of modification the mesocoxae of 
the present fly are likely to belong to the last category because 
of the presence of the large membranous area around the base 
of the coxa, which may be homologous with the meron, and not 
due to the undivided nature of the coxae. Thus the mesocoxae 
of this fly must be quite different in constitution from the pro- 
coxae, although they resemble closely in appearance those of the 

The metacoxa of the Diptera, as a whole, is not subdivided 
superficially, while in the related orders, Trichoptera, Mecoptera, 
and Lepidoptera, all the metacoxae are distinctly subdivided into 
the two sclerites as well as in the mesocoxae. Whether the me- 
tacoxa here is undivided as in the procoxae or represented by 
the eucoxa alone as in the mesocoxa of certain dipterous insects, 
is quite problematic. 

The trochantin is rarely found in the Diptera, and it is com- 
monly said that the legs of Diptera are directly articulated to 
the thorax without the intermediate sclerites. I have pointed 
out an exception in the case of the Tipulidae where this sclerite 
is present in a highly reduced semichitinous state. In the 
subject of this paper the sclerite is lacking. 

The trochanters of the dipterous legs generally are very short, 
small, and somewhat triangular in outline, no matter how the 
legs and coxae may be elongated, as in the crane flies and fungus 
gnats. In the present fly all the trochanters are very long and 
cylindrical. Moreover, each trochanter is firmly fused with the 
proximal subdivision of the femur along the immovable femasu- 
ture. This fused part moves as if it were one segment, being 

"According to Young (1921) the female of a certain midaid fly, Mu 
das clavatus, is exceptional in being provided with the independent meron. 

184 The Philippine Journal of Science 1935 

demarked by the trochacoria and the secondary membranous 
ring of the femur at each end, and is not provided with muscles, 
which suggests the absence of movement at the femasuture. 
This immovable trochafemoral union is widely found in various 
insects, even in the more primitive orders, as in the Thysanura, 
but it is most prevalent in the higher orders, although in all of 
these cases the trochanters and femora are fused into one seg- 
ment, while in Nymphomyia the membranous ring of the femur 
subdivides it into the proximal part, which consists of the tro- 
chanter and the proximal end of the femur, and the distal part, 
which represents the large distal region of the femur. 

The dipterous femur usually is simply elongated and somewhat 
flattened in its cephalocaudal diameter and subdivided femora, 
like those in Nymphomyia, are rarely found among the Diptera. 
They are only known in the highly specialized parasitic flies, 
the Nycteribiidse, and in the Deuterophlebiidse (female) .^^ 

On the tibiae of this fly are also subdivisions similar to those 
in the femora, and they are equally rare. This pseudosegmental 
nature of the tibise is shown in spider flies as the tibial rings 
and more definite constrictions are found in many flies and 
caddice-fly larvse.^^ 

The above discussion on the subdivisions of the femora and 
tibise leads to the conclusion that the pseudosegmental feature 
in the present insect may result from the mechanical necessity 

"In the latter case Pulikovsky, Trans. Ent. Soc. London (1924) 45-62, 
stated, after examining a female adult, which was extracted from the pu- 
pal skin, that the trochanter consists of two joints, but I confirmed by ob- 
servations on females, both the pupal and emerged adult, that the so-called 
second trochanter is nothing but a part of the femur, which is partially 
constricted or subdivided on the dorsal side by a semicircular membranous 
area. At this membrane the leg in the pupal skin is bent sharply as if it 
were a true segment, but in the fully extended leg of the free-living imago 
it reveals a structure similar to that of Nymphomyia. Thus, unisegmental 
trochanters are the rule among the Diptera, and the true plurisegmental 
nature, which is found in certain limited groups (for example, Odonata 
and Symphyta and certain parasitic Hymenoptera) , is never found among 
the Diptera. 

^* Among the Arachnoidea there is a large segment, the patella, between 
the femur and tibia, and in a few insects this segment is found as a small 
intersegmental sclerite, as I have stated in the case of a certain crane 
fly. Research on the homology between the patella and the proximal sub- 
division of the tibia and also on that between the second trochanter and 
the proximal subdivision of the femur, at least in dipterous legs, failed 
to reveal any relation between the two, respectively. 

^^' 2 Tokunaga: A Nymphomyiid Fly 185 

of providing flexibility to the segment by secondary division or 

The accessory structures of the tarsal claws, such as the 
pulvilli and empodium, are of considerable importance for clas- 
sificatory purposes, since they are constantly present in dipterous 
insects. In the Cyclorrhapha the two padlike pulvilli are always 
present, although the padlike empodium sometimes is reduced 
to a bristlelike form. In the Brachycera i» of the superf amily 
Eremochseta, as a rule, both the pulvilli and empodium are pad- 
like in structure, while the rest of this series is provided with 
two padlike pulvilli and one bristlelike empodium. In the Nema- 
tocera certain families, such as the Bibionidae and Rhyphidse, 
are usually provided with well-developed pulvilli and empodium 
similar to the Eremochseta, but in the great majority of this 
series two pulvilli are completely wanting or vestigial, although 
a single padlike empodium is retained. Thus, both the absence 
of the two pulvilli and the presence of the single padlike em- 
podium of the Nymphomyiidae offer strong evidence that this 
fly belongs in the series Nematocera. 


The abdominal segments of the insect are divided into three 
groups; namely, pregenital, genital, and postgenital. The first 
term is applied to the cephalic seven abdominal segments, which 
contain most of the visceral organs. The second consists of the 
eighth and ninth segments, which are highly modified generally 
on account of the adaptive structures for copulation and oviposi- 
tion. The last term is used for the segments beyond the ninth, 
which are greatly reduced and fused with each other. 


The pregenital segments (common abdominal segments, vis- 
ceral segments of Snodgrass) of the Diptera are modified and 
reduced in various degrees according to the taxonomic groups. 
In the higher groups they are reduced to five or less, while in the 
nematocerous group there are seven, which are typical and show 

^^Verall, G. H., British Flies. London 5 (1909) 14-47, arranged the 
brachycerous flies as follows: Ereraochasta (Stratiomyiidae, Acanthome- 
ridae, Leptidae, Tabanidae, Nemestrinidse, and Cyrtidae), Tromoptera (Bom- 
byliidae and Therevidae), Dermatina (Scenopinidae and Mydaidae), Ener- 
gopoda (ApioceridaB and Asilidse) Microphona (Empidae and Dolichopo- 
didaB), Acroptera (Lonchopteridae), and Hypocera (Phoridae). 

186 The Philippine Journal of Science 1935 

no special differentiation in either sex, as in the species under 
consideration. In the brachycerous group they show transi- 
tional forms between the above two groups, and the number of 
the nonmodified abdominal segments is variable according to 
taxonomic division and sex, but tolerably constant within the 
limits of most of the families. Among the brachycerous in- 
sects the hypothetical number of the nonmodified segments is 
shown in a few families, such as the Leptidse, Asilidse, Empidae, 
Scenopidae, and Mydaidae. In the majority of other families 
one, two, or more posterior abdominal segments are reduced 
and modified into the telescopic or concealed genital segments. 
The significance of the reduction and modification of the ab- 
dominal segments is apt to be overlooked, but it must have some 
phylogenetic importance when taken in connection with the 
study of the genital segments. 

The seven segments of this region as a rule are simple in 
structure, and in the Nematocera differ but little from one 
another either in shape or size. The first segment is more 
subject to modification than is any other. The first segment of 
the Diptera is somewhat different in size from the adjacent 
abdominal segment; and, as Young has pointed out, there is a 
tendency for the first tergum, as well as the corresponding 
sternum, to decrease in relative size. Usually the first segment 
is smaller than the second, and this is best exhibited in the 
higher Diptera. In the lower Diptera, such as the Chironomo- 
idea, Tipuloidea, Psychodoidea, and Culicoidea, the first segment 
is usually very large but not larger than the second, while in 
the present fly the first segment, especially its tergum, is un- 
usually large and larger than the second. Moreover, this large 
segment of Nymphomyia, as in the general Orthorrhapha, is 
never due to the fusion with the second segment, which is quite 
common throughout the Cyclorrhapha.^^ 

* There has been reported a secondary suture of the first abdominal ter- 
gum named the "adventitious suture'* by Young (1921) who stated that 
this suture "was found in all the species examined among both the Calyp- 
tratsB and the AcalyptrataB. Generally speaking this suture is less de- 
veloped in the calyptrate than in the acalyptrate muscids, etc." But I am 
not yet certain whether or not the presence or absence of this suture is an 
important definite mark of distinction between the Cyclorrhapha and the 
Orthorrhapha, since in certain nematocerous insects, e. g. chironomid flies — 
similar thickenings are found, although their homology with the adventi- 
tious suture remains to be elucidated (Tokunaga, 1932). 


Tokunaga: A Nymphomyiid Fly 


In homologizing the sclerites or divisions throughout the ab- 
dommal segments, the position of the spiracles is an invaluable 
guide, but unfortunately in Nymphomym the abdominal spiracles 
are completely lacking, and the sclerites are also greatly reduced 
into a common membrane, thus obliterating the usual abdominal 
divisions. In the case of this fly the lateral groove (text fig 5 
x-^x) is the only mark by v^hich the divisions of the abdominal 
segments can be determined. This groove is distinct in the 
larvae of holometabolous insects, 
extending along each side of the 
body throughout the abdomen 
and the thorax below the line 
of the spiracles. It is named 
the "dorso-pleuraP' or "tergo- 
pleural groove" by Snodgrass 
(1931), the "sterno-pleural su- 
ture" by Craighead (1916), the 
"ventrolateral suture" by Bov- 
ing (1914), and the "pleural 
suture" by Hopkins (1909) . As 
shown by the names applied, the 
morphologic significance of this 
groove is differently stated by 
each of these writers, but I in- 
tend to employ the name "tergo- 
pleural groove" given by Snod- 
grass, because of the following 
facts: (a) In the thoracic region 
this groove extends between the 
paratergite, which is the lateral 
sclerite of the tergum or the 

allied region, and the subcoxa, which is homologous with the 
pleuron; and (6) on the genital segments this groove is between 
the paratergite or allied region and the common limb base of the 
stylus and gonapophysis, which is also homologous with the 
thoracic pleuron. 

In the present fly this tergopleural groove (text fig. 2, Y-Y) 
extends longitudinally below the lateral tubercle and the lateral 
transparent pattern, and is most distinctly shown in the state 
of contraction of the abdominal muscles. Besides this groove 
there is another very shallow groove (text fig. 2, Z-Z) , which 
also extends longitudinally above the lateral tubercle or below 

Fio. 5. Diagram of the theoretical abdo- 
minal segment. (Modified from Snod- 
grass, 1931.) as, Antecostal suture; gB, 
gonapophysis; Iot, limb base or pleuron; 
p, precosta ; pa, paratergite ; S, sternum ; 
8, spiracle; ay, stylus; T, tergum; x-z, 
tergopleural line. 

188 The Philippine Journal of Science issfi 

the dorsal group of the small tubercles. This shallow groove 
may be regarded as being derived from the dorsal attachments 
of the transversal lateral muscles, somewhat similar to the 
tergopleural groove. For this reason this groove is thought 
to be homologous with the paratergal suture. Thus we can 
distinguish in the abdominal segment of this fly the tergum 
(s. lat.) and the pleurosternum by the tergopleural groove. The 
tergum is subdivisible into the tergum (s. str.) and the para- 
tergite by the paratergal groove, but the sternum and the 
pleuron cannot be delimited as they form a completely uniform 

The tergum of Nymphomyia is not different in its essential 
structures from that of insects in general, being provided with 
the intersegmental membrane, precosta, and antecosta. 

The pleurosternal region of this fly also differs but little in 
the essential structures from that of insects in general. The 
absence of distinct chitinization and differentiation of precosta 
and antecosta are the only results of reduction and modification, 
a condition not very significant since it is very prevalent in 
the higher orders, not only among the Diptera. Moreover, the 
distinction between the sclerotization of the pleuron 21 and the 
sternum (primary sternum) is known in the lower Apterygota 
and immature forms alone (text fig. 5), whereas in the imagines 
of the higher Holometabola these two sclerites are completely 
fused with each other, or the pleuron is completely reduced to 
a membrane, or both are reduced to a common membrane, as 
in the present fly. In the latter case, therefore, the so-called 
**sternum" or "sternite" is either composed of the united pleuron 
and primary sternum or of the last sclerite alone. 

The structure of the paratergal region of this fly is peculiar 
in the absence of the abdominal spiracle and the presence of 
the scleritelike tuberculate patch (text fig. 2, It). The non- 
spiracle or apneustic condition of the dipterous abdomen has 
been reported by Edwards and Kemper for D enter ophlehia 
mirabilis and Psychoda phalaenoides, and this condition is very 
peculiar for the adult Diptera. The paratergal region of the 
Diptera generally is completely membranous where the spiracle 
is found. Sometimes the tergal plate is not divided into the 

**The pleuron of the abdominal segment is known as the "limb base" 
in the Apterygota and the campodeiform njrmphs, and as the "subcoxa" in 
certain holometabolous cruciform larvse (Coleoptera, Hymenoptera, and 

56, 2 

Tokunaga: A Nymphomyiid Fl/y IgQ 

tergum and paratergite and extends so far downward as to 
include the spiracles as seen in certain Cyclorrhapha (Tachi- 
nidse, Muscidse, Sarcophagidse, Anthomyiidse, etc.). While in 
the present form the scleritelike patch is present in the position 
where the spiracle would be expected. This external differen- 
tiation 22 of the paratergal sclerotization is very unusual in the 
adult Holometabola, since it is found only in the immature forms 
of certain Coleoptera, Lepidoptera, and Hymenoptera. The 
origin of this patch, whether it is derived from the preexistent 
paratergal sclerite of the larval stage or from the secondary 
chitinization in the imaginal stage, remains to be investigated. 


In the higher orders of insects, especially the Diptera, the 
genital and postgenital divisions are very obscure due to the 
secondary fusion, reduction, and modification for copulation 
and oviposition. The number of the genital segments is 
usually considered to be two, but that of the postgenital seg- 
ments has received diverse interpretations by different author- 
ities. There have been known two representative theories. 
One is the eleven-segment theory, in which the abdominal seg- 
ments are counted as nine somites, together with the two post- 
genital somites, the tenth and eleventh (telson) somites, of 
which the cerci are the appendages of the tenth abdominal 
somite. The other is the twelve-segment theory, which is main- 
tained embryologically by Heymons, Wheeler, Nelson, and others, 
and morphologically by Snodgrass, Imms, MacGillivray, and 
others, who recognize three postgenital somites, the tenth, elev- 
enth, and twelfth, the cerci properly belonging to the eleventh 
somite, and the twelfth somite being sometimes known as the 
telson. On the whole, the balance of opinion inclines towards 
the latter theory. In the Diptera this primitive theoretical con- 
dition is not known to occur, since the posterior-most two, three, 
or more segments are usually fused with each other and become 
membranous. In the female of the Tipulidse, however, the ninth, 
tenth, and cercus-bearing eleventh, abdominal segments may be 
distinguished as I reported in 1930, and the females of the 
Panorpidse (Mecoptera) show similar abdominal segmentation. 

Based on the positions of the genital aperture, anus, para- 
meres, cerci, and cercarise in the female and of the clasping 

"Internally the paratergal region is usually characterized in almost all 
insects by the presence of a special group of the paratergal muscles. 

190 The Philippine Journal of Science 1935 

organs in the male, the following segmentation of the terminal 
region, as illustrated in text fig. 6, may be suggested for the 
present species. 

The anterior genital segment of the female fly differs very 
much in structure from other Diptera in the presence of the 
characteristic lateral extension. This lateral extension (Plate 

4, figs. 14 to 16; Plate 5, figs. 17 to 19, pj) is the specially 
modified tergopleural fold. Thus, the caudal extensions of this 
fold have no relation to the true appendages of limb bases, such 
as the styli and gonapophyses (text fig. 5, sy and gs) , although 
they superficially assume a position proper for the typical genital 
appendages. The latter are greatly reduced, as in the general 
Diptera, and represented only by thickened parameres (text 
fig. 6, gsi) guarding the aperture of the gonad in the inner- 
most part of the genital chamber. Among the Diptera the 
t3rpical gonapophyses and limb bases of the eighth abdominal 
segment are shown only in the females of the Tipulidse; in all 
other Nematocera they are highly reduced, as in Nymphomyia 
or represented by small, paired, membranous lobes, which are 
known as the octavalvae (Tokunaga, 1930 and 1932). In the 
Brachycera and Athericera these structures are completely 
atrophied. In the Nymphomyiidse the aperture of the gonad 
is typically situated between the eighth and ninth sterna, as in 
the general nematocerous insects, and has not migrated caudad 
as in the higher groups having telescopic ovipositors. Another 
modification of the eighth segment is the longitudinal ental 
thickening of the sternum. This thickening (Plate 4, fig. 16, 
ec) is developed secondarily for the attachment of the muscles 
related to the female genital organs. 

In the male the eighth segment is as peculiar as in the female 
in the presence of the characteristic paratergal structures (Plate 

5, figs. 17 to 19, pj) , although they are different in shape from 
the latter. The essential sexual differences are the complete 
atrophy of the gonapophyses and the related appendages and 
the absence of both the sternal antecostal thickening and the 
secondary ental chitinization found in the female. Other fun- 
damental structures of this segment are closely similar to those 
of the pregenital segment, as in the case of the female. 

The ental thickenings of the genital and postgenital segment 
develop secondarily and sporadically in various regions in widely 
varying shapes among different groups (families, genera, or 
species) for mechanical reasons, and between opposite sexes 
according to the variation of the hypopygial structures. This 


Tokunaga: A Nymphomyiid Fly 

Pig. 6. Diagrams illustrating the concept of the structural plan of the terminal abdominal 
sesrments adopted in this paper. 1, Dorsal aspect, male; £, ventral aspect, male; S, 
ventral aspect of the postgenital segments, male ; 4, dorsal aspect, female ; 5, ventral 
aspect, female, a, Anus; aa, cedeagus; at, antecosta; ep, epiproct; ga, genital aperture: 
ga, gonapophysis or paramere ; hn, limb base, coxite or cercaria ; pb, pons basalis ; pi, 
periproct ; ptt, pleuron ; pt, paratergite ; S, sternum ; a, stylus or cercus ; aa, lami subanalis ; 
«c, segmacoria ; su, lami supra-analis ; t, tergum ; vaa, ventral wall of genital chamber. 

192 The Philippine Journal of Science 1935 

is best exhibited in the hypopygia of chironomid flies such as 
Clunio, TanytarstLs, and Chironomus. 

In both sexes the paratergal projections are devoid of both 
the special musculature and the mechanism of articulation, in- 
dicating that they are merely a local elongation of the paratergal 
region and not the regular appendage. In the male sex the 
paratergal projection is movable dorsoventrad, but this is due 
to the contraction of the tergosternal muscles, which extend 
from the common basal plate (paratergite proper) of the pro- 
jection to the sternum and are not due to the independent move- 
ment of the projection by itself. 

The eighth segment of the male, as a whole, is not normally 
provided with genital appendages, no matter how highly it may 
be modified as in the concealed or retractile genital form of the 
higher Diptera. The sedeagus {aa) is found superficially just 
caudad of the eighth sternum in this insect, but it properly be- 
longs to the next caudal segment as indicated by a deep mem- 
branous furrow formed by the segmacoria between it and the 

In the great majority of dipterous insects, the genital ninth 
and postgenital somites in both sexes are highly modified from 
their original features and fused with each other, forming the 
so-called "terminal ninth abdominal segment." The tenth and 
twelfth segments, especially, are most unstable and reducible, 
and it is very difficult to identify them. In this highly modified 
and reduced condition the important marks are the persistent 
parameres, cerci, and various clasping organs, which show the 
location of their own segments. Even when the dorsum is 
represented by a completely fused sclerotization, they indicate 
some segmental divisions on the venter. 

The dorsum of the terminal segment in the present female 
is uniformly thickened and there is no trace of segmentation, 
but the presence of paired cerci, which closely arise from the 
lateral sides of the dorsum, shows that the eleventh tergum 
is concerned with the formation of this dorsum. The ninth 
tergum may be considered either as reduced to a common mem- 
brane, forming a large segmacoria, or fused with the posterior 
terga, forming a common dorsum. In the present fly the latter 
seems to be the case, since the ninth tergum always remains 
last when the reduction of these segments occurs, and the 
eleventh tergum never remains alone without the coexistence 

^®' ^ Tokunaga: A Nymphomyiid Fly 193 

of the chitinization of the ninth tergum. Thus, the dorsum of 
the terminal segment is regarded as formed by the complete 
fusion of the ninth, tenth, and eleventh terga. 

The venter of the female terminal segment is generally mem- 
branous and divisible into two regions by a transversal thick- 
ening, which is the second paramere (the first parameres are 
the paired sclerites of the eighth segment found at the genital 
aperture at the end of the eighth sternum). Thus, the dorsal 
wall of the genital chamber, cephalad of the second paramere, 
is the proper ninth sternum. The caudal region, caudad of 
this paramere, where the cercariae, cerci, and blunt paraprocts 
are found, is formed by the fusion of the tenth and eleventh 
sterna and periproct. The periproct may be regarded as the 
remnant of the highly reduced twelfth somite. Thus, the venter 
of the female terminal segment is divisible into two or three 
regions; namely, the ninth sternum and the common venter of 
the postgenital sterna or the twelfth is further separated from 
the latter. 

The terminal segment of the male carries large paired clasping 
organs, cerci, and unpaired sedeagus. The presence of these 
structures shows that this segment is formed by the fusion 
of the ninth and postgenital segments. The dorsum of this 
segment does not show segmentation, but the lateral sides pos- 
sess slight constrictions that suggest the demarcation between 
the clasper-bearing ninth and the cercus-bearing eleventh seg- 
ments. There can be little doubt, therefore, that the middorsal 
area is formed by the fusion of the proper ninth, tenth, and 
eleventh terga. The small area between the bases of the large 
cerci may be the remnant of the twelfth tergum, known as the 
lami supra-analis (su). 

The venter of the male terminal segment is more distinctly 
divided into two regions than on the dorsum by the presence 
of the large sedeagus and the distinct cercariae. The sedeagus 
is formed by the fusion of the median penis and the paired 
lateral parameres, which are homologous with the gonapophyses 
of the ninth segment. Thus, in the present fly it seems to be 
the caudal projection of the eighth sternum, but it properly 
belongs to the ninth segment, as already stated. The large 
pubescent area caudad of the sedeagus, where the cercariae are 
located, may be the fused venter of the proper sterna of the 
tenth and eleventh segments. The most caudal small narrow 

289853 7 

194 The Philippine Journal of Science i^ss 

area between the bases of the cerci corresponds to the lami 
supra-analis (su) of the periproct (the twelfth somite) and is 
known as the lami subanalis (sa) . 

Usually the tenth abdominal somite of the Diptera is highly 
or completely atrophied in both sexes, but its location is fre- 
quently shown by the accessory genital apparatus apparently 
homologous with the gonapophyses and styli of the tenth somite. 
These apparatus generally are found concealed by the large 
clasping organs of the male. In certain Rhyphidae, Culicidse, 
Chironomidse, etc., they are paired and situated close to the bases 
of the coxites of the clasping organs belonging to the ninth 
somite. In other families, such as the Blepharoceridae, Psycho- 
didas, etc., they become unpaired, median, valvelike projections 
closely applied on the dorsal side of the sedeagus. 

In the great majority of the Diptera the cerci superficially 
arise from the tenth somite and thereby the cercus-bearing seg- 
ment is counted as the tenth, but this is due to the reduction 
of the true tenth somite. The cerci must belong to the eleventh 
abdominal somite, however, because of the retention of the 
appendages properly belonging to the tenth somite. 

For these reasons, the twelve-segment theory is considered 
duly applicable to the dipterous abdominal segmentation as well 
as to that of the generalized insect. 

To recapitulate the main points in the homology of the 
appendages of the terminal abdominal somites of the Nym- 
phomyiidse, the following may be stated. The lateral projections 
of the eighth somite in both sexes have no relation to the true 
pleural appendages, being paratergal in origin. The true ap- 
pendages of the eighth somite remain as the first parameres in 
the female, while those of the male are completely atrophied. 
Those of the ninth fiomite of the female are also reduced to the 
second parameres. In the male they are highly developed, the 
limb bases and styli forming the large paired clasping organs 
and the gonapophyses taking part in the formation of the large 
sedeagus. The appendages of the tenth somite are completely 
atrophied in both sexes. The styli and limb bases of the eleventh 
somite in both sexes are homologous with the paired cerci and 
the paired cercariae, respectively. The twelfth somite of each 
sex is reduced to the periproct (text fig. 6). 

The hypopygial structures show wide divergence among the 
Diptera, and their phylogenetic affinities are very difficult to 
ascertain. Taken in connection with other structures, the fol- 

56» 2 Tokunaga: A Nymphomyiid Fly 195 

lowing features as a whole are considered important in estab- 
lishing the affinity of this fly to the nematocerous group: The 
nonmodified pregenital segments, the nontelescopic or non- 
concealed hypopygium of each sex, the comparatively simple 
structures of the male hypopygium, the large elongated cerci 
of the male located on the caudal end of the abdomen, the large 
elongated coxites situated laterally, the large exposed aedeagus, 
the absence of the chitinized ejaculatory filament, and the typical 
position of the genital apertures in both sexes. 


Although the family Nymphomyiidse possesses the brachy- 
cerous antennae, it is considered as belonging to the Nema- 
tocera, on the strength of the following distinctive characters: 
Elongated head capsule, large occipital foramen, cylindrically 
elongated thorax, nonangulated mesopleural suture, nondifferen- 
tiated abdominal segments, exposed and comparatively simple 
structures of the genital segments and their appendages, large 
and prominent cerci of both sexes, absence of pulvilli, presence 
of empodium, and the nematocerous wing venation. 

This family seems to stand on a comparatively low level in the 
Nematocera, because of the following thoracic structures : ^^ 
The large pronotum, small pleuron, isolated and large sternum, 
and elongated coxse. Other structures, large cerci, large cox- 
ites, and styles which are typically arranged, are also thought 
to be rather primitive characters for the Diptera. This con- 
sideration is supported anatomically by the presence of the 
nonconcentrated independent eight ganglia in the abdomen, as 
will be reported in a later contribution. However, when the 
antennal structures, the wing venation, reduced metanotum, 
fused terminal segments of the abdomen, modified eighth ab- 
dominal segment of both sexes, etc., are taken into considera- 
tion, this insect is not so low in position as to be compared 
with the protonematocerous insects or with the archaic ancestral 
Diptera. The weight of the various extraordinary characters 
of this insect, then, leads us to the conclusion that this species 
is a representative of an extremely specialized family isolated 
from all other groups of the Diptera. 

*»The thoracic sclerotization has a vastly more important significance 
for the consideration of phylogenetic affinity of the Diptera than other 
parts and various appendages, because of its comparative stability, as 
suggested by Crampton (1925 and 1926) and Tokunaga (1932). 

196 The Philippine Journal of Science i»35 


The conclusions derived from morphological study of the scle- 
rotization of the nymphomyiid fly may be itemized as follows: 

A. The chief characteristic modification of the head and mouth 
parts has occurred along the line of reduction of various struc- 
tures and in the direction of cephalocaudal elongation of the 
head capsule as a whole. 

1. The cephalocaudal arrangement of various organs, such as 
the antennae, compound eyes, and ocelli, and also the progna- 
thous type of the head, are intimately related to the elongation 
of the head capsule. 

2. The head capsule consists of the vertex, f rontoclypeus, occi- 
put, and probably the antennarise, stipites, and theca compactly 
fused together. 

3. The occipital foramen is unusually large, which may be due 
to the prognathous head capsule. 

4. The following homology of the peculiar projections of the 
head capsule may be shown : 

Snoutlike projection. Frontoclypeal projection. 

Visorlike projection. Cranial projection. 

Labiumlike projection. Projection of the theca? or stipes? 

15. The following structures of the fixed parts of the head are 
.atrc^hied : 




Other articular processes, 



Their tentorin®. 


Distal part of the stem. 


Frontocljrpeal, etc. 

The proximal part of the epicranial stem alone remains as 
the middorsal suture on the dorsal part of the head. 

6. The great majority of the mouth parts is highly reduced. 
"The pharynx directly opens into the peculiar chitinized pocket- 

^®' 2 Tokunaga: A Nymphomyiid Fly I97 

like mouth concavity. Developmental states of the trophic or- 
gans are shown as follows: 

Movable parts. 

Paraglossae? (papilliform projection). 


Maxillas and appendages. 
Labium and appendages. 



Postpharynx. (Esophageal pump. 
Salivary pump. 





Salivary sclerites. 

7. The antennas are of the brachycerous type and provided 
with the characteristic intersegmental sensillae as in the larval 
antennae of the Diptera. The constitution of the antenna is 
shown as follows: 


Antennaria (probably fused with the adjacent sclerite of the 

head capsule). 
Antacoria (very small). 

Scape (pyriform). 
Pedicel (spherical). 
Flagellum (3-segmented). 

First segment (large and spoon-shaped). 
Second segment (small and ring-shaped). 
Intersegmental sensillaB (clavate and three in number). 
Third segment (small and needle-shaped). 

8. The compound eyes are of the holoptic type and charac- 
teristic in being contiguous on the ventral side while widely 
separated from one another on the dorsal side. They consist of 
the normal oculatae and biconvex facets. 

9. The ocelli are very large, laterally located, and two in num- 
ber. They may be different in origin from the primitive ocelli 

198 The Philippine Journal of Science 1935 

or primary ocelli, but more probably are derived from the lateral 
ocelli or larval eyespots. 

10. The muscles are also reduced, and only twelve different 
muscles are contained within the head capsule. 

B. All the sclerites of the cervix are completely degenerated. 
The cervix is broad but short and becomes as a whole uniform- 
ly membranous. 

C. The thorax is highly different from the dipterous thorax in 
general, both in characteristic modification and in the rudimen- 
tal condition of its sclerotization. 

1. The notum of the prothorax is large and secondarily di- 
vided into the antenotum and postnotum, which are completely 
subdivided into paired lateral halves on the dorsomeson by the 
cephalic projection of the mesoprsescutum and become located on 
the lateral side. 

2. The mesonotum consists of the following sclerites : 







Katapleurotergites ? (parasternites ) . 

The elongation of the thorax, on the dorsal side, is mainly 
due to the elongated scutoprsescutum and postscutellum (mesa- 
scutella). The mesascutella deeply intrudes caudoentad and 
forms a large phragma, which is one of the two endoskeletons 
well developed in the thorax. The parascutella occupies almost 
the caudal half of the entire latus of the thorax, and its anomalous 
development affects the position of the mesoepimeral sclerites 
and also the relative position of the middle and hind legs. 

3. The sclerites of the metanotum are completely reduced into 
a uniform membrane, being infolded between the mesothoracic 
sclerite and the cephalic margin of the first abdominal antecosta. 

4. The development of the pleural sclerites is very poor. This 
may be partially due to the rudimental and membranous condi- 
tion of the pleuron itself and also to the effect of the intrusion 
of the parascutella. 

56.2 Tokunaga: A Nymphomyiid Fly 199 

5. The pleuron is very small, being represented only by the 
episternal chitinization from which the procoxacoila is derived. 
The epimeral sclerites are completely atrophied. 

6. The mesopleuron is comparatively small and its chitiniza- 
tion is very low in development. The pleural suture is not 
sharply angulated but slightly undulated. The episternal suture 
is obscure. The epimeral suture is completely obliterated by the 
cephalic intrusion of the parascutella along this suture. The 
displacement of the epimeral sclerites is chiefly due to the 
anomalous expansion of the parascutella. 

7. The metapleuron is represented only by the epimeral chiti- 
nization, completely losing the episternal sclerites. 

8. The coxacoilse of the meso- and metathoracic segments are 
not developed. 

9. The sternal sclerites are fully developed and loosely ar- 
ranged, as shown in the thorax of certain immature forms or 
certain lower insects. Viewed from the evolutional standpoint 
of the thoracic sclerotization of the Diptera, these sternal fea- 
tures, taken in connection with the rudimental condition of the 
pleuron, indicate that the thorax is in the primitive state of 
development. The elongation of the thorax, on the ventral side, 
is mainly related to the large sternal sclerites and their loose 

10. The constitution of the sternal sclerites is shown as fol- 



Undivided primary sternum. 

Fused laterosternites. 

Independent presternum. 

Membranous basisternum. 

Elongated sternellum. 

Fused precoxal bridges. 

Secondary chitinization? (parasternites). 


Fused laterosternites. 

Fused pre- and postcoxal bridges. 


The Philippine Journal of Science 


Each sternum is provided with a pair of sternacoilse which 
are derived from the laterosternites for the articulation of the 
coxae. The postcoxal bridges of the mesosternum and the pres- 
ternum of the metasternum are atrophied. The unusual lateral 
position of the procoxal articulation is related to the broad pres- 
ternum and the vestigial propleuron. 

11. The thoracic endoskeletons are generally very low in de- 
velopment. Their relation to the external marks is shown as 
follows : 

Posi tion. 


External mark. 

Development, etc. 



Prostemum J 



Mesosternum .. 




Phragmae «-._- 




Obscure, along the pleural suture. 



Very large, along the cephalic 
margin of the praescutum. 

Small, along the cephalic margin 
of the scutulis. 

Small, along the laterocaudal 
margin of the scutulis. 

Very large, along the caudal mar- 
gin of the mesascutella, fused 
with the metaprephragma. 

Narrow, along the caudal margin 
of the parascutella. 

Obscure, along the pleural suture. 

Narrow, along the caudal margin 
of the presternum. 

Small, along the transternal su- 

Obscure, along the midventral su- 
ture, originated from the furca. 

Obscure, along the pleural suture. 

United with the postphragma of 
the mesothorax. 


Obscure, along the cephalic mar- 
gin of the basisternosternellum. 

Small, along the midventral su- 
ture, originated from the furca. 

Pieuradema ..-_- 


Furca - 








Lateral phragma 



Caudal ohracma _«_-__ 

Pieuradema - 

Pleurademina — 

Transversal phragma 

Furca_ - 


Longitudinal phragma__ 






Prephragma - 

Phragm a 


Loniritudinal Dhra&ma 

D. The legs are all slender but small for the size of the insect. 
The forelegs are the longest and the hind legs are the shortest. 

56»2 Tokunaga: A Nymphomyiid Fly 201 

1. The segmentation of the legs is shown as follows: 

Morphological segmentation. 



Coxal rim 

Basicoatal suture. 
Coxa proper 

Mechanical segmentation. 



First segment. 
Articulated at the trochacoria between the artes and coilae. 

Second segment. 


Proximal region _ _ 

Membranous ring. 
Distal region 

Third segment. 

Articulated at the tibiacoria between the artes and coil». 


Proximal region 

Membranous ring. 
Distal region 

Fourth segment. 

Fifth segment. 

Articulated at the tarsacoria between the trodia and flexis. 
Five tarsal segments I Sixth to tenth segments. 

2. The coxse are cylindrically elongated. The mesocoxa con- 
sists of the eucoxa alone and the meron is reduced. The coxal 
rim is not provided with special apodemal discs. 

3. Both the trochantin and patella are atrophied. 

4. The terminal structures of the leg are as follows: 










E. The wings are highly characteristic in form, due mainly 
to the great reduction of various structures and to the wide 
modification of the marginal structures. 

1. The wings are very narrow and cuneiform, due to the 
reduction of the alula, squama, and anal lobe. 

202 The Philippine Journal of Science 1935 

2. The marginal setse are extremely developed. They are 
thought to serve to increase the soaring coefficient of the insect 
in the air. 

3. The well-developed ambient vein is related to the develop- 
ment of the marginal fringe and to the reduction of the true 

4. The wings are deciduous along a definite line at the wing 
basis just distad of the venellse. 

5. The venation is very obscure. The state of development 
of the veins is shown as follows : 


Costa and subcosta (coalesced with each other). 
Radius ( 2-branched ) . 

First radial branch. 

Radial sector (unbranched) . 
Media (unbranched?). 

First cubitus and second cubitus (coalesced with each other). 

Anal veins. 
Secondary veins. 

6. The basal sclerites and processes of the wings for the ar- 
ticulation are generally reduced in number and size. The state 
of their development is shown as follows: 



Subalifera (L-shaped and large). 
Prealifera (small). 
Medalifera (oval and obscure). 
Postalifera (oval and large). 

Pleuralif era proper. 
Ental discs. 


Sigmoidea? (obscure). 
Terminalia (small). 
Submedia? (obscure). 
Ental discs. 


^•' ^ Tokunaga: A Nymphomyiid Fly 203 

Pteraliae — Continued. 


Costalis (large). 
Radiales (small). 
Analis (small). 


Medalaria (minute). 
Caudalaria? (obscure). 
Scutalaria (minute). 

Other structures. 

Tegula (vestigial). 

Spiralis (comparatively large), 

Ponta (rod-shaped and thickly chitinized). 

F. The halter es are normal in structure. The great major- 
ity of the basal sclerites and the articular processes are atro- 
phied. The structures retained are the following: 


Scabellum (small). 
Petiole ( elongated ) . 

C+Sc4-R (cephalic margin; obscure). 
M+Cu+A (caudal margin; obscure). 
Capitellum ( spoon-shaped ) . 
Basal sclerite. 

Tegula (vestigial). 

G. The pregenital segments are seven in number, and they are 
not distinctly differentiated in both sexes, being subequal in 
shape and structure to each other. 

1. The divisions and subdivisions of these segments are as 
follows : 

Pregenital segment. 
Tergum (s. lat.). 

Tergum (s. str.). 

Paratergal groove. 

Tergopleural groove, 

2. The abdominal spiracles are all atrophied. The various 
cuticular structures and their arrangement are shown in text 
fig. 1. 

204 The Philippine Journal of Science 1935 

H. The segmentation of the terminal abdominal region and the 
relation between it and various appendages and projections are 
shown in text fig. 5. 

1. The constitution and homology of the female genital seg- 
ments are shown as follows: 

Eighth segment. 

Tergum (s. str.). 

One projection. 
Two tubercles. 


Stylus ( atrophied ) . 
Gonapophysis (first paramere). 
Limb base (obscure). 


Thickened area (sternum proper). 
Ental ridge (secondary). 

Membranous area (ventral wall of the genital chamber). 
Ninth segment. 

Tergum (fused with that of the postgenital segments). 


Stylus (atrophied). 
Gonapophysis (second paramere). 
Limb base (atrophied). 
Sternum (dorsal wall of the genital chamber). 

The genital aperture is found between the eighth and ninth 
sterna, being guarded by the first parameres. 

2. Those of the male genital segments are as follows: 

Eighth segment. 

Tergum (s. str.). 

Basal plate (paratergite proper). 
Two projections. 
Three tubercles. 

Appendages ( atrophied ) . 
Limb base (obscure). 
Sternum (fused with the pleuron forming the pleurosteiHum). 

66' 2 Tokunaga: A Nymphomyiid Fly 205 

Ninth segment. 

Tergum (fused with the common terminal tergum). 


Stylus (retained). 

Gonapophysis (paramere of the aedeagus). 
Limb base (coxite). 
Sternum (pons basalis). 

Thus, the sedeagus consists of the parameres, median penis, 
pons basalis, and ejaculatory duct. 

3. The postgenital segments are completely fused in both 
sexes. Their constitution in the female is as follows : 

Tenth segment. 

Tergum (fused with the common terminal tergum). 
Pleuron (obscure). 
Sternum (obscure). 
Eleventh segment. 

Tergum (fused with the common terminal tergum). 


Stylus (prominent cercus). 
Gonapophysis ( atrophied ) . 
Limb base (cercaria). 
Sternum (obscure). 
Twelfth segment (periproct). 

Tergum (lami supra-analis) . 
Sternum (lami subanalis). 

4. The constitution of the postgenital segments in the male 
Is as follows : 

Tenth segment (obscure). 
Eleventh segment. 

Tergum (obscure). 


Stylus (prominent cercus). 
Gonapophysis ( atrophied ) . 
Limb base (cercaria). 
Sternum (obscure). 
Twelfth segment (periproct). 
Tergum (lami supra-analis). 
Sternum (lami subanalis). 

I. The Nymphomyiidse are considered to be a comparatively 
low nematocerous family and so specialized as to be quite iso- 
lated phylogenetically from all other dipterous groups. 

206 The Philippine Journal of Science 1935 

Amans, p. C. 

Comparisons des organes du vol dans la serie animale. Ann. Sci. Nat, 
VI 19 (1885) 9-222. 
Berless, a. 

Gli Insetti, loro organi25zazione, sviluppo, abitudini e rapporti coiruomo, 
Soc. Edit. Libraria, Milano (1909). 
BoviNG, A. G. 

On the abdominal structure of certain beetle larvae of the campodei- 
form type. A study of the relation between the structure of the 
integument and the muscles. Proc. Ent. Soc. Washington 16 (1914) 
Cholodkowsky, N. 

Die Embryonalentwicklung von Phyllodromia (Blatta) germanica. 
Mem. Acad. Sci. St. Petersburg VII 38 (1891) 1-120. 
Christophers, S. R., and P. J. Barraud,. 

The development of the male and female hypopygium of Phlebotomus. 
Ind. Journ. Med. Res. 13 (1926) 853-870. 

The Wings of Insects. Ithaca, New York (1918). 
An Introduction to Entomology. Ithaca, New York (1924). 
CoMSTOCK, J. H., and C. Koch. 

The skeleton of the head of insects. Am. Nat. 26 (1902) 13-45. 
Collins, C. W. 

Dispersion of gipsy-moth larvae by the wind. Bull. U. S. Dept. Agr. 
273 (1915) 1-23. 
Craighead, F. C. 

The determination of the abdominal and thoracic areas of the ceram- 
bycid larvae as based on a study of the muscles. Proc. Ent. Soc 
Washington 18 (1916) 129-142. 
Crampton, G. C. 

A contribution to the comparative morphology of the thoracic scle- 
rites of insects. Proc. Acad. Nat. Sci. Philadelphia 1909 (1909) 
The ground plan of a typical thoracic segment in winged insects. 

Zool. Anz. 44 (1914) 56-67. 
Notes on the thoracic sclerites of winged insects. Ent. News 25 

(1914) 15-25. 
The nature of the veracervix or neck region in insects. Ann. Ent. 

Soc. Am. 10 (1916) 187-197. 
A phylogenetic study of the terminal abdominal segments and ap- 
pendages in some female apterygotan and lower pterygotan in- 
sects. Journ. N. Y. Ent. Soc. 25 (1917) 225-237. 
Thoracic sclerites of grasshopper Dissosteira Carolina. Ann. Ent 
Soc. Am. 11 (1918) 320-347. 

^^' 2 Tokunaga: A Nymphomyiid Fly £07 

Crampton, G. C. — Continued. 

A phylogenetic study of the terminal abdominal structures and geni- 
talia of male Apterygota, ephemerids, Odonata, Plecoptera, Neu- 
roptera, Orthoptera, and their allies. Bull. Brooklyn Ent Soc. 
13 (1918) 49-68. 

The genitalia and terminal abdominal structures of male Neuroptera 
and Mecoptera with notes on the Psocidae, Diptera, and Trichoptera 
Psyche 25 (1918) 47-59. 

Notes on the ancestry of Diptera, Hemiptera, and other insects re- 
lated to Neuroptera. Trans. Ent. Soc. London 1919 (1919) 93- 

A phylogenetic study of the mesothoracic terga and wing bases in 
Hymenoptera, Neuroptera, Mecoptera, Diptera, Trichoptera, and Le- 
pidoptera. Psyche 26 (1919) 58-64. 

A comparison of the external anatomy of the lower Lepidoptera and 
Trichoptera. Psyche 27 (1920) 23-44. 

A comparison of the genitalia of male Hymenoptera, Mecoptera, Neu- 
roptera, Diptera, Trichoptera, Lepidoptera, Homoptera, and Sterp- 
siptera, with those of lower insects. Psyche 27 (1920) 34-45. 

Remarks on the basic plan of the terminal abdominal structures of 
the males of winged insects. Canad. Ent. 52 (1920) 178-183. 

A comparison of the terminal abdominal structures of insects and 
Crustacea. Ent. News 32 (1921) 257-264. 

The sclerites of the head, and the mouth-parts of certain immature 
and adult insects. Ann. Ent. Soc. Am. 14 (1921) 65-100. 

The phylogeny and classification of insects. Journ. Ent. Zool. 16 
(1924) 33-47. 

Remarks on the phylogeny and interrelationships of nematocerous 
Diptera. Psyche 31 (1924) 238-242. 

A phylogenetic study of the thoracic sclerites of the non-tipuloid ne- 
matocerous Diptera. Ann. Ent. Soc. Am. 18 (1925) 49-67. 

A phylogenetic study of the thoracic sclerites of the psychodoid Dip- 
tera, with remarks on the interrelationships of the Nematocera. 
Ent. News 37 (1926) 33-39 and 65-70. 

A comparison of the neck and prothoracic sclerites throughout the 
orders of insects from the standpoint of phylogeny. Trans. Am. 
Ent. Soc. 52 (1926) 199-248. 

The thoracic sclerites and wing bases of the roach, Periplanata amer- 
icana, and the basal structures of the wings of insects. Psyche 
24 (1927) 59-72. 

The evolution of insects, chilopods, diplopods, Crustacea, and other 
arthropods indicated by a study of the head capsule. Canad. Ent. 
60 (1928) 129-141. 
Crampton, G. C, and W. H. Hasey. 

The basal sclerites of the leg in insects. Zool. Jahrb. Anat. 39 
(1915) 1-26. 

208 The Philippine Journal of Science 1935 

Bmsss, L. 

Das larvale Muskelsystem und die Entwicklung der imaginalen Flug- 
muskulatur von Psychoda alternata Say. Zeit. Morphol. Okol. Tiere 
11 (1928) 182-228. 
Edwards, P. W. 

The nomenclature of the parts of male hypopygium of Diptera Nema- 
tocera, with special reference to mosquitoes. Ann. Trop. Med. 
Paras. 14 (1920) 23-40. 
Deuterophlebia mirabilis, gen. et sp. nov., a remarkable dipterous in- 
sect from Kashmir. Ann. & Mag. Nat. Hist. 9 (1922) 379-387. 
Hammond, A. 

Thorax of the blow-fly. Journ. Linn. Soc. London 15 (1881) 9-31, 
Handursch, a. 

Die fossilen Insekten und die Phylogenie der rezenten Formen. Leip- 
zig (1908). 
Heberdy, K. E. 

Zur Entwicklungsgeschichte, vergleichenden Anatomie und Physio- 
logie der weiblichen Geschlechtsausfiihrwege der Insekten. Zeit. 
Morphol. 5kol. Tiere 22 (1931) 416-586. 
Hesse, R. 

Untersuchungen uber die Organe der Lichtempfindung bei niederen 
Thieren, VII. Von den arthropoden Augen. Zeit. wiss. Z06I, 70 
(1901) 347-473. 
Heymons, R. 

Die Segmentierung des Insektenkorpers. Abhandl. Akad. Wiss. Ber- 
lin 1895 (1895) 1-39. 
Zur Morphologie der Abdominalanhange bei den Insekten. Morphol. 

Jahrb. 24 (1896) 178-204. 
tJber die abdominalen Korperanhange der Insekten. Biol. Centralb. 

16 (1896) 854-864. 

Entwicklungsgeschichtliche Untersuchungen an Lepisma saccharima L. 

Zeit. Wiss. Zool. 62 (1897) 583-631. 
Der morphologische Bau des Insektenabdomens. Zool. Centralb. 6 

(1899) 537-556. 
Die Hinterleibsanhange der Libellen und ihrer Larven. Ann. K. 

Naturhist. Hofmuseums. Wien 19 (1904) 21-58. 
Hopkins, A. D. 

The genus Dendroctonus. U. S. Dept. Agr. Bur. Ent. Tech. Ser. No. 

17 (1909) 1-164. 
Imms, a. D. 

A General Textbook of Entomology. London (1925). 
Kellogg, V. L. 

The mouth-parts of the nematocerous Diptera. Psyche 8 (1899) 303- 
306, 327-330, 846-348, 355-359, 363-365. 
Kemper, H. 

Morphogenetische Untersuchung liber das Tracheensystem von Psy- 
choda phalaenoides (Diptera). Dissertation, Miinster (1925). 

^^'^ Tokunaga: A Nymphomyiid Fly 209 

MacGillivray, a. B. 

External Insect-anatomy. Urbana, Illinois (1923). 
Matheson, R. 

The mosquitoes of North America. Springfield, Illinois (1929), 


Studies on the fruit-flies of Japan. I. Japanese orange-fly. Bull Imp 
Cent. Agr. Exp. Stat. 2 (1919) 87-160. 
Nelson, J. A. 

The Embryology of the Honey-Bee. Princeton Univ. (1915). 
Morphology of the honey-bee larva. Journ. Agr. Res. 28 (1924) 1167- 
Newell, A. G. 

The comparative morphology of the genitalia of insects. Ann. Ent. 
Soc. Am. 11 (1918) 109-142. 
Peterson, A. 

The head-capsule and mouth-parts of Diptera. 111. Biol. Monog 3 
(1916) 1-62. 
Prochnow, 0. 

Mechanik des Insektenfluges. Handb. Ent. (Schroder) 1 (1924) 534- 
Samtleben, B. 

Anatomie und Histologie der Abdominal- und Thorax-muskulatur von 
Stechmiickenlarven. Zeit. wiss. Zool. 134 (1929) 180-269. 
Saunders, L. B. 

Some marine insects of the Pacific coast of Canada. Ann. Ent. Soc. 
Am. 21 (1928) 521-545. 
Snodgrass, R. E. 

The terminal abdominal segments of female Tipulid«. Journ. N. Y. 

Ent. Soc. 11 (1903) 177-183. 
The thorax of insects and the articulation of the wing. Proc. U. 

S. Nat Mus. 38 (1909) 511-595. 
Anatomy and metamorphosis of apple maggot, Rhagoletis pomonella 

Walsh. Journ. Agr. Res. 28 (1924) 1-36. 
Morphology and mechanism of the insect thorax. Smith. Misc. Coll. 

80 (1927) 1-108. 
Morphology and evolution of the insect head and its appendages. 

Smiths. Misc. Coll. 81 (1928) 1-158. 
The thoracic mechanism of a grasshopper, and its appendages. Smiths. 

Misc. Coll. 82 (1929) 1-109. 
Morphology of the insect abdomen. Part I. General structure of 
the abdomen and its appendages. Smiths. Misc. Coll. 85 (1931) 
Tillyard, R. J. 

The Insects of Australia and New Zealand. Sydney (1926). 

289853 8 

210 The Philippine Journal of Science 


The morphological and biological studies on a new marine crane fly, 
Limonia (Dicranomyia) monostromia, from Japan. Mem. Coll. 
Agr. Kyoto Imp. Univ. No. 10 (1930) 1-93. 
Morphological and biological studies on a new marine chironomid fly, 
Pontomyia pacifica, from Japan. Part I. Mem. Coll. Agr. Kyoto 
Imp. Univ. No. 19 (1932) 1-56. 
A remarkable dipterous insect from Japan, Nymphomyia alba gen. et 

sp. nov. Ann. Zool. Jap. 13 (1932) 559-569. 
A phylogenetic study of the thoracic sclerites of the Deuterophlebiidae. 
Rep. Jap. Ass. Adv. Sci. 8 (1932) 428-434. 
Walker, E. M. 

The terminal structures of orthopteroid insects: a phylogenetic study, 
Part II. Ann. Ent. Soc. Am. 15 (1922) 1-76. 
Weber, H. 

Lehrbuch der Entomologie. Jena (1933). 
Weinland, E. 

tJber die Schwinger (Halteren) der Dipteren. Zeit. wiss. Zool. 51 
(1890) 55-166. 
Wheeler, W. M. 

The embryology of Blatta germanica and Doryphora decimlineata. 

Journ. Morphol. 3 (1889) 291-386. 
A contribution to insect embryology. Journ. Morphol. 8 (1893) 1-160. 
Young, B. P. 

Attachment of the abdomen to the thorax in Diptera. Mem. Cornell 
Univ. Agr. Exp. Stat. Mem. 44 (1921) 255-306. 

The anatomy of the head and mouth-parts of Orthoptera and Euplexop- 
tera. Journ. Morphol. 33 (1920) 251-307. 


Nymphomyia alba Tokunaga 
Plate 1. Head and Antenna 

Fig. 1. Lateral aspect of head. 

2. Dorsal aspect of head. 

3. Ventral aspect of head. 

4. Dorsal aspect of dextral antenna. 

ac, Antacoria. 

6a;, Basipharynx. 

ce, Compound eye. 

es, Epicranial stem. 

fl, Segment of flagellum. 

iSf Intersegmental sensilla. 

la, Papilliform projection 

Ip, Labiumlike projection, 
wc, Mouth cavity. 
o. Occipital margin. 

Plate 2. 
Fig. 5. Lateral aspect. 

6. Dorsal aspect. 

7. Ventral aspect. 

ad, Abdominal segment. 

an, Antepronotum. 

6s, Basisternum. 

c, Coxafossa. 

cc. Cervix or cervacoria. 

CO, Coxacoila. 

em, Epimeron. 

en, Ental invagination or 

caudal phragma. 
et, Episternum. 
h, Haltere. 
la, Lateral phragma of 

m, Mesascutella. 
ma, Medalaria. 
mc, Mesocoria. 
md, Middorsal suture. 
ms, Midventral suture, 
n, Notepimeron. 
p, Pleural suture. 
pa, Paraphragma. 
pc, Parascutulis. 

oc, Ocellus. 

of, Occipital foramen. 

ol, Oculata. 

op, CEsophageal pump. 

p, Pedicel. 

sn, Snoutlike projection. 

sp, Scape. 

ty Toothlike projection. 

vr. Visorlike projection. 

vt, Vertex. 


pi, Postscutellum. 

pn, Postpronotum. 

pp, Propleuron. 

prs, Presternum. 

ps, Parascutella. 

pst, Prosternum. 

pt, Parasternite. 

pv, Pseudosutural fovea. 

r, Rotaxis. 

8, Scutalaria. 

sa. Supra-alar seta. 

sc, Sternacoila. 

se, Sternepimeron. 

si, Scutulis. 

si, Spiralis. 

sm, Sternellum. 

sps, Scutopraescutum. 

ss, Secondary (transsternal) suture. 

st, Transversal stripe. 

su, Scutellum. 

W8, Wing base. 



The Philippine Journal of Science 


Plate 3. Lbsgs and Haltere 

Fig. 8. Foreleg. 
9. Middle leg. 
10. Hind leg. 

c, Claw. 

en, Calcanea. 
ex. Coxa. 

d, Distal subdivision. 

e, Empodium. 
/, Femur. 

fs, Femasuture. 

gh, Glenkhocker. 

mc, Basicostal suture of 

mf, Membranous ring of 

mt, Membranous ring of 


p, Proximal subdivision. 

sn, Sensilla. 

t, Trochanter. 

ta, Trochartis. 

th, Tibia. 

tha, Tibiartis. 

thCf Tibiacoila. 

tc, Trochacoila. 

to, Trochacoria. 

tp, Tarsal spine. 

ts, Tarsal segment. 

tu, Tubercula. 

11. Sinistral haltere, lateral aspect. 

hsr. Basal swelling of ro- ps, Parascutella. 

ca, Capitellum. 
em, Epimeron. 
et, Episternum. 
in, Invagination between 
thorax and abdomen, 
m, Mesascutella. 

ptl, Petiole. 

r, Rotaxis. 

sad, Sternal side of abdomen. 

sc, Scabellum. 

sp. Porelike puncture (spiracle?). 

tad, Tergal side of abdomen. 

te, Tegula. 

Plate 4. Wing and Terminal Segments of Female 
Fig. 12. Wing. 

am, Anal margin. 

av, Ambient vein. 

C, Costa, 

cm, Costal margin. 

Cu, Cubitus. 

M, Media. 

13. Wing base, ventrolateral aspect 

om, Outer margin. 
ps, Placoid sensilla. 
R, Radius. 
Rs, Radial sector. 
Sc, Subcosta. 

as, Analis. 
cs, Costal is. 
mxi, Medalaria. 
ml, Medalifera. 
n, Notepimeron. 
np, Notepisternum. 
p, Pleural suture. 
pal, Prealifera. 
po, Ponta. 
pol, Postalifera. 
pr, Subalifera (pleurali- 

ps, Parascutella. 

r, Rotaxis. 

ra, Radialis. 

sa, Supra-alar seta. 

se, Sternepimeron. 

si, Spiralis. 

spy Porelike puncture (spiracle?). 

su, Scutellum. 

t, Terminalia. 

te, Tegula. 

W8, Wing base. 


Tokunaga: A Nymphomyiid Fly 


14. Terminal segments of female, dorsal aspect. 

15. Terminal segments of female, lateral aspect. 

16. Terminal segments of female, ventral aspect. 

a, Anus. 

at, Antecosta. 

cc, Cercaria. 

ce, Cercus. 

ec, Ental thickening. 

ep, Epiproct. 

gcy Genital chamber. 

OS, Octasternum. 

ot, Octatergum. 

pj, Paratergal projection. 

pp, Paraproct. 

pt, Paratergite. 

sa, Lami subanalis. 

sc, Segmacoria. 

su, Lami supra-analis. 

t, Tubercle. 

Plate 5. Terminal Segments op Male 

Fig. 17. Lateral aspect. 

18. Dorsal aspect. 

19. Ventral aspect. 

a, Anus. 

aa, u^deagus. 

at, Antecosta. 

cc, Cercaria. 

ce, Cercus. 

ex, Coxite. 

h, Harpe. 

if, Intersegmental 

OS, Octasternum. 
ot, Octatergum. 
p, Penis. 


pa, Paramere. 

pb, Ponta basalis. 

pj, Paratergal projection. 

pp, Paraproct. 

pt, Paratergite. 

s, Style. 

sa, Lami subanalis. 

sc, Segmacoria. 

su, Lami supra-analis, 

t, Tubercle. 

te, Telson. 


Pig. 1. Head of Nymphomyia alba Tokunaga. Cross sections of the basi- 
pharynx; a, through the attachments of the anterior dilators; b, 
through those of the posterior dilators; and, c, of the posterior 
region before the oesophagus. 

bx, Basipharynx. 

en. Connective nerve. 

fg, Frontal ganglion. 

oe, (Esophagus. 

ol, Oculata. 

op, CEsophageal pump. 

p. Pedicel. 

sbg, Subcesophageal ganglion. 

sd. Salivary duct. 

sg, Salivary gland. 

sip. Salivary pump. 

sp, Scape. 

spg, Supraoesophageal ganglion. 

2. Diagrammatic arrangement of the cuticular structures of the ab- 
dominal segment. 

ac, Antecosta. 

It, Lateral tubercle. 

tb, Tubercule. 

tp, Transparent pattern. 
X-X, Tergoparatergal line. 
Y-Y, Tergopleural line. 


The Philippine Journal of Science 

3. Hypothetical definitive sternum. (Modified from Snodgrass, 1931.) 

acx, Precoxal bridge. 
Bs, Basisternum. 
ex, Coxa. 
fs, Furcal suture. 
Ls, Laterosternite. 
pcx, Postcoxal bridge. 
PSf Presternum. 

4. Hypothetical venation. 

Sc, Subcosta. 

Ri, First branch of radius. 

Rs, Radial sector. 

sa, Sternartis. 

sc, Sternacoila. 

SI, Sternellum. 

spn, Spinafurca. 

sps, Secondary presternal suture. 

Ss, Spinasternum. 

M, Media. 
Cu, Cubitus. 

5. Diagram of the theoretical abdominal segment. (Modified from 

Snodgrass, 1931.) 

as, Antecostal suture. 

gs, Gonapophysis. 

Im, Limb base or pleu- 

p, Precosta. 
pa, Paratergite. 

S, Sternum. 

s, Spiracle. 

sy, Stylus. 

T, Tergum. 

{C-aj, Tergopleural line. 

6. Diagrams illustrating the concept of the structural plan of the 
terminal abdominal segments adopted in this paper, 1, dorsal as- 
pect, male; 2, ventral aspect, male; S, ventral aspect of the post- 
genital segments, male; 4, dorsal aspect, female; 5, ventral as- 
pect, female. 

pn, Pleuron. 
pt, Paratergite. 
S, Sternum, 
s, Stylus, or cercus. 
sa, Lami subanalis. 
sc, Segmacoria. 
su, Lami supra-analis. 
T, Tergum. 

vaa. Ventral wall of genital cham- 

a, Anus. 

aa, i^deagus. 

at, Antecosta. 

ep, Epiproct. 

ga. Genital aperture. 

gs, Gonapophysis, or pa- 

Im, Limb base, coxite, 

or cercaria. 
pb. Pons basalis. 
pi, Periproct. 

Tokunaga: a Nymphomyiid Fly.] 

[Philip. Journ. Sci., 56, No. 2. 


Tokunaga: a Nymphomyhd Fly.] 

[Philip. Journ. Scl, 56, No. 2. 



Tokunaga: a Nymphomyiid Fly.] 

[Philip. Journ. Scl, 56, No. 2. 


Tokunaga: a Nymphomyud Fly, 

Philip. Journ. Scl, 56, No. 2. 


Tokunaga: a Nymfhomyiid Fly.] 

[Philip. Joukn. Sci., 56, No. 2. 





By Claro Martin and Heraclio R. Montalban 
Of the Fish and Game Administration, Bureau of Science, Manila 


The present paper contains a review of the species of the 
family Parapercidse known to inhabit Philippine waters. It is 
based on specimens in the collection of the Fish and Game Ad- 
ministration, Bureau of Science. 


Body elongate, sybcylindrical, posteriorly compressed, covered 
with small ctenoid scales ; head a little depressed ; eyes more or 
less lateral, close together; mouth moderate or denticulate; in- 
teropercle usually not entire; opercle with a spine; subopercle 
either blunt or pointed ; gill membranes united, free from isth- 
mus; branchiostegals 6; pseudobranchise present; dorsal fin long, 
usually continuous; anal similar to soft dorsal; ventrals thora- 
cic, I~5; air bladder absent; pyloric caeca few. 

Small fishes found in reefs and along shores, sometimes in 
deeper waters, from the Red Sea and the east coast of Africa 
through the seas of India to Polynesia. 

According to the scheme of classification of Jordan the Para- 
percidsB are closely related to the Mugiloididse on one side and 
to the Pteropsaridse on the other. Species of these two families 
have so far been neither collected nor reported from the Phil- 

Genus PARAPERCIS Bleeker 
Parapercis Bleeker, Ned. Tijdschr. Dierk. 1 (1863) 236. 

Body cylindrical, rather elongate; cleft of mouth oblique; teeth 
in villif orm bands with small curved canines ; no teeth on pala- 
tines ; teeth on vomer ; dorsal divided, spinous portion with four 
or five spines. 

Key to the Philippine species of Parapercis Bleeker, 

a\ Spinous dorsal with five spines. 
b\ Caudal with black dots; no caudal blotch. 


216 The Philippine Journal of Science 1935 

c \ Suprascapular region without a blotch. 

Parapercis cylindrica (Bloch). 
c*. Suprascapular region with a white-and-black-ringed brown spot. 

Parapercis tetracantha (Lacepede). 
b^. Caudal blotch present. 

c\ Blotch on caudal large, black; sides of body with three to seven 

ocelli Parapercis hexophthalma (Cuvier and Valenciennes). 

c^ Blotch on caudal white; cheeks with oblique, narrow bands. 

Parapercis dorsonebulosa sp. nov. 
a^. Spinous dorsal with four spines. 

b ^. Suprascapular region with a white-edged ocellus, head with black 

dots Parapercis clathrata Ogilby. 

b ^, Suprascapular region not ocellated ; head with brown blotches. 

Parapercis montillai sp. nov. 

PARAPERCIS CYLINDRICA (Bloch). Plate 1, fig. 1. 

Scimna cylindrica Bloch, Ausl. Fische 6 (1792) 42, pi. 299, fig. 1. 
Percis cylindrica Cuvier and Valenciennes, Hist. Nat. Poiss. 4**. 3 

(1829) 199; Bleeker, Nat. Tijdschr. Ned. Ind. 2 (1863) 235; 

GUNTHER, Cat. Brit. Mus. 2 (1860) 239. 
Parapercis cylindrica Jordan and Richardson, Bull. Bur. Fisheries 

27 [(1907) 1908] 281; Weber, Siboga Exp., Fische (1913) 519; 

Fowler, Mem. P. Bishop Mus. 10 (1928) 424. 

Dorsal V, 21; anal 18 or 19; scales on lateral line 49 or 50; 
between lateral line and origin of dorsal 4 ; between lateral line 
and origin of anal 17 or 18. 

Body elongate, moderately compressed, its depth 4 to 4.6 in 
length ; depth of caudal peduncle 2.5 to 3 in head ; dorsal profile 
almost as strongly arched as ventral outline, but a little more 
curved; head small, pointed, 3.4 to 3.7 in length; snout acutely 
rounded at tip, 2.5 to 3.3 in head ; eye small, moderately high, 
almost equidistant from tip and edge of opercle, 3.4 to 3.9; 
interorbital flat, wide, 8.8 to 11.4; mouth small, oblique; lips not 
broad; premaxillaries not very protractile; maxillary almost 
entirely concealed by preorbital, extending to, or a little behind, 
anterior border of eye; upper jaw with an outer series of en- 
larged teeth which are coarser in front, with one short curved 
canine on each side and a band of villiform teeth inside; lower 
jaw with an outer series of coarse teeth in front, with a long 
canine on each side at the termination of the series, and an 
inner villiform band in front, continuing posteriorly to a single 
series of enlarged ones; villiform teeth on vomer. A spine at 
upper angle of opercle; interopercle and subopercle terminate 
in a spine ; that of former sometimes blunt. 

A rather deep notch between the spinous and soft dorsals; 
third and fourth spines longest, as long as snout; or slightly 

56, 2 Martin and Montalban: Philippine Parapercidm 217 

longer; anterior dorsal rays as long as distance from tip of snout 
to hind margin of pupil; anal inserted below fifth dorsal ray; 
pectoral small, rounded, 4.5 to 4.9 ; ventral slender, reaching as 
far as sixth anal ray, 3.1 to 3.7; caudal rounded or slightly so, 
with upper lobe pointed but very slightly produced. 

Grdund color in alcohol grayish white to yellowish, nine brown- 
ish to blackish vertical bands on each side of body, the middle 
ones spindle-shaped ; a short dark-edged band on snout from its 
tip to eye where it joins another one crossing middle of eye long- 
itudinally or slightly obliquely; top of head with a coarse network 
of very pale blue, inclosing brownish spaces; a vertical brown 
band from lower part of eye across cheek; a narrower band of 
same color running obliquely across jugular region, meeting its 
kind ventrally; a blackish blotch with a narrow strip of white 
above at basal half of spinous dorsal; pectoral and ventral, 
respectively, yellowish brown and brownish gray in life; soft 
dorsal with yellow edge and basal black spots; dusky areas at 
anterior portion gradually become defined as small spots pos- 
teriorly; anal with many small dusky spots sometimes fused to 
form vertical bands; caudal brownish in life, spotted with dark 
blackish brown, larger at base, with white edge and with a 
submarginal dusky border. 

This species is seemingly the commonest of the family. It is 
represented in the collection by specimens ranging in length 
from 48.5 to 135 millimeters collected from different localities 
from Polillo southward to Sitankai. 

PARAPERCIS TETRACANTHA (Lacepede). Plate 1, fi^. 2. 

Labrus tetracanthus Lacep!3de, Hist. Nat. Poiss. 4°, 3 (1802) 428, 

473, [not Percis tetracanthus Bleeker, Nat. Tijdschr. Ned. Ind. 

4 (1853) 458, which is P. quadrispinosus (Weber)]. 
Percis cancellata Cuvier and Vai*enciennes, Hist, Nat. Poiss. 4**. 3 

(1829) 200; Bleeker, Nat. Tijdschr. Ned. Ind. 9 (1855) 501; 

Gunther, Cat. Brit. Mus. 2 (1860) 240. 
Parapercis tetracanthus Jordan and Richardson, Bull. U. S. Bur. 

Fish. 27 [(1907) 1908] 281; Weber, Siboga Exp., Fische 65 (1913) 

Parapercis tetracantha Ogilby, Proc. Roy. Soc. Queensl. 23 (1910) 

40. Not described. 

Dorsal V, 21 ; anal 18 ; scales on lateral line 59 to 60 ; between 
lateral line and origin of dorsal 8 ; between lateral line and origin 
of anal 19 to 22. 

Body elongate, rather cylindrical, its .depth 5.9 to 6.2 in length; 
caudal peduncle compressed, its depth 3 to 3.3 in length of head; 

218 The Philippine Journal of Science isss 

dorsal outline of head and body slightly more arched than ven- 
tral; head rather depressed with moderately bulging cheeks, 

3.3 to 3.5 in length, snout regular, 2.8 to 3 in head; eye large, 
much advanced forward, its upper margin slightly projecting 
above head, 4.6 to 4.9; interorbital narrow, 9.2 to 11; mouth 
moderately large, oblique; lips broad; lower jaw projecting; pre- 
maxillaries protractile ; posterior portion of maxillary partly ex- 
posed, extending a little beyond anterior border of eye; upper 
jaw with an outer series of coarse teeth with several long ones 
in front and on the sides, and an inner villiform band narrow- 
ing posteriorly; lower jaw with an outer series of coarse enlarged 
teeth, four in front and three at each side directed posteriorly, 
and an inner villiform band not continuous behind ; teeth also on 
vomer; preopercle entire; a strong, rather flat spine on upper 
angle of opercle. 

Membrane uniting spinous and soft dorsals low; third and 
fourth spine longest, as long as eye or about equal to it; an- 
terior dorsal rays as long as distance from snout to about mid- 
dle of pupil and inserted below fifth dorsal ray ; pectoral short, 
broadly rounded, its length 5.4 to 5.7 in body; ventral pointed, 

4.4 to 5.2; caudal rounded, with projecting upper and lower 
lobes, the former a little longer than the latter, its length 5.2 
to 5.7. 

Scales behind occiput, on cheeks, breast, and abdomen smooth ; 
finely ctenoid all over the rest of body and base and upper and 
lower lobes of caudal; snout, sides of head, and occiput naked. 

Color in alcohol dark brown, lighter below on each side; 
three series of light blotches with slightly darker centers ; eight 
blotches along each side of back, comprising the first series, 
each continuous with the corresponding blotch on the other side ; 
second series along middle of body consists of nine alternating 
with those above, the third corresponding with the second, each 
with white edge, narrow above and wider below across the 
belly, and continuing with that of the other side; a black dot on 
each side of tip of snout; a long blackish blotch on each side 
of upper lip; lower lip and chin brown; a broad brown blotch 
on cheeks below eyes ; a large white-and-black-ringed brown spot 
on suprascapular region; a black vertical blotch below base of 
pectoral; two pairs of small black blotches on belly, sometimes 
connected; a larger black blotch below base of pectoral; mem- 
branous portion of first dorsal whitish with some shades of 
brownish; soft dorsal transparent with three series of black 

66.2 Martin and Montalban: Philippine Parapercidse 219 

spots, those at base fewer and larger; anal with whitish edge 
and dusky submarginal border ; caudal with a number of dusky- 
spots with black centers ; base of upper lobe of caudal with small 
dusky brown blotch. 

Here described from four specimens, 170 to 200 millimeters 
long, from the following localities : Mindoro, Mindoro Province, 
Calapan. Romblon, Romblon Province, Romblon. Mindanao, 
Cotabato Province, south coast 

PARAPERCIS HEXOPHTHALMA (Curier and Valenciennes). Plate 2, fiir. 1. 

Percis cylindrica Ruppell, Atl. Reise Nord. Afrika, Fische (1828) 18, 

pi. 5, fig. 2 (not Scimna cylindrica Bloch, which is P. cylindrica 

Cuvier and Valenciennes). 
Percis hexophihalma Cuvier and Valenciennes, Hist. Nat. Poiss. 4 *. 

3 (1829) 202; GUNTHER, Cat. Brit. Mus. 2 (I860) 239; Playfair 

and GuNTHER, Fish. Zanzibar (1866) 68; Day, Fishes of India 4\ 

(1878) 263. 
Percis polyophthalma Cuvier and Valenciennes, Hist. Nat. Poiss. 

4*. 3 (1829) 203; Playfair and Gunther, Fish. Zanzibar (1866) 

68; Klunzinger, Fisch. Roth. Meer. 1 (1870) 816. 
Percis caudimaculata Ruppell, Neue Wirbelthiere, Fische (1835-1840) 

98; Bleeker, Verb. Bat. Gen. 22 (1849) 54. 
Parapercis hexophthalma Jordan and Snyder, Proc. U. S. Nat. Mus, 

24 (1902) 466; Evermann and Seale, Bull. U. S. Bur. Fish. 26 

(1906) 103; Weber, Siboga Exp., Fische 65 (1913) 518; Barnard, 

Ann. So. African Mus. 21 (1925-27) 422, pi. 19, fig. 1; Fowler, 

Mem. Bernice P. Bishop Mus. 10 (1928) 424. 
? Parapercis tetracanthus Jordan and Richardson, Bull. U. S. Bur. 

Fish. 27 (1908) 281. 

Dorsal V, 21; anal 18; scales on lateral line 60 to 61; between 
lateral line and origin of dorsal 8, between lateral line and 
origin of anal 19 to 23. 

Body elongate, subcylindrical, its depth 5.3 to 5.6 in length; 
depth of compressed caudal peduncle 2.8 to 3 in head ; dorsal out- 
line at head region more arched than ventral; head pointed, 
3.4 to 3.5 in length; snout quite long, 2.5 to 2.7 in head; eye 
large, nearer tip of snout than edge of opercle, located high, 
the upper margin projecting above outline of head, 4.2 to 4.6; 
interorbital narrow, 8.4 to 11; mouth moderate, oblique, lips 
broad; jaws not quite equal; premaxillaries protractile; maxil- 
lary concealed by preorbital, reaching a little beyond anterior 
border of eye; upper jaw with an outer series of conical teeth, 
with nine or ten in front enlarged, curved and caninelike, and 
with an inner wide band of small ones narrowing posteriorly; 
lower jaw with eight teeth in front and three enlarged teeth at 

220 The Philippine Journal of Science ins 

middle of outer series, and an inner big patch of small ones ; a 
small patch of fine teeth on the vomer; preopercle entire with a 
strong rather flat spine at posterior angle. 

Membrane connecting spinous and soft dorsals high; third and 
fourth spine about equal, almost as long as eye diameter or 
slightly longer; rays much higher than spines, anterior ones 
equal to distance from snout to pupil; anal inserted below fifth 
dorsal ray; pectoral rounded, 4.7 to 4.9 in body length; ventrals 
pointed, 4 to 4.5 ; caudal rounded with upper pointed lobe slightly 

Scales behind occiput, on cheeks, and on breast smooth; on 
the rest of body and on caudal fin finely ctenoid; snout, inter- 
orbital space, and occiput naked. 

Color in alcohol yellowish to grayish brown with irregularly 
scattered blackish spots above, whitish below; round blackish 
dots all over head ; on sides seven to eight white, somewhat rec- 
tangular, spaces with one or several blackish spots inside, and a 
row of blackish spots below; three to seven black ocelli, each 
surrounded by a narrow white ring, the first one small, below 
base of pectoral, the last three along side of abdomen being per- 
sistent; in larger specimens the first four being reduced to un- 
defined dusky spots; a black spot at base of spinous dorsal; 
three rows of black to brownish spots along soft dorsal, a nar- 
row dusky strip edged with white forming a row near the tip of 
each ray; anal with narrow white tip, an indistinct dusky sub- 
margin, and a narrow dusky band at base, and with a row of 
blackish to dusky spots; caudal with a large black blotch, with 
a number of black spots and small white blotches behind and 

Here described from four specimens, 185.5 to 206 millimeters, 
from the following localities: Luzon, Batangas Province, Ha- 
milo. Tablas, Romblon Province, Tablas. Balabac, Palawan 
Province, Balabac. 

We have also examined a specimen from Japan. None of the 
specimens was found to have "the oblique brown lines radiat- 
ing from the lower part of the eye, over the opercles" men- 
tioned by Playfair and Giinther. 

PARAPERCIS DORSONEBULOSA sp. nor. Plate 2, fig. 2. 

Parapercis hexophthalmus Jordan and Seale, Bull. U. S. Bur. Fish. 
26 [(1906) 1907] 46 (non Bleeker). 

56,2 Martin and Montalban: Philippine Parapercidx 2i2X 

Dorsal V, 20 to 21; anal 18; scales on lateral line 60; between 
lateral line and origin of dorsal 7; between lateral line and 
origin of anal 17 to 18. 

Body elongate, subcylindrical, its depth 5.6 to 5.8 in length; 
caudal peduncle compressed, its depth 3.1 to 3.4 in head ; dorsal 
and ventral profiles almost equal; head moderate, 3.3 to 3.5, 
with cheeks full but not bulging; snout gently sloping to spa- 
tulate tip, 2.6 to 2.8 in head; eyes large, 4.1 to 4.22; interorbi- 
tal flat, 12.6 to 19 ; mouth large, oblique ; lips broad ; premaxilla- 
ries protractile; maxillary entirely concealed under preorbital, 
its length extending beyond anterior border of eye; lower jaw 
projecting; upper jaw with an outer series of enlarged teeth 
coarser in front and posteriorly curved canines on each side; 
vomer with a narrow band of villiform teeth; opercle with a 
rather strong spine at upper angle; subopercle terminating in 
a blunt flat spine. 

Dorsal continuous, inserted above and slightly behind base of 
pectoral ; third or fourth dorsal spine longest, equal to diameter 
of eye or slightly less ; anterior dorsal rays as long as distance 
from snout to behind front margin of pupil or sligthly beyond ; 
anal inserted below fifth dorsal ray; pectoral 4.6 to 5.5 in body 
length; ventrals short, pointed, 4 to 4.8; caudal rounded, with 
a projecting pointed upper lobe. 

Scales somewhat coarsely ctenoid, present all over body and 
caudal except snout and top and sides of head, smooth scales 
on cheeks, behind occiput, on breast, and anterior portion of 

Ground color in alcohol brownish to yellowish white and 
brownish above in old specimens. Specimens of three months 
preservation in alcohol, back dusky with a number of scales here 
and there splashed with black; a white longitudinal band on 
side from above base of pectoral to base of caudal; eight to 
nine dusky crossbars starting from the band and disappearing 
below distant from median ventral line; eight to nine oblique, 
narrow, white lines across cheeks and anterior portion of oper- 
cle; dorsal with three rows of black dots, uppermost near edge 
of fin; one on anal; pectoral and ventral white to brownish; 
caudal with black spots and a large white blotch on posterior 
two-thirds of middle rays. 

222 The Philippine Journal of Science 1935 

TjTpe : No. 31252, in the Bureau of Science collection, 140 milli- 
meters long, collected from Catbalogan, Samar, June, 1934, by 
Juan Julayco. Three cotypes, one obtained from the same place, 
and two from Balabac Island, Palawan Province. 

Dorsonebvlosa, cloudy back, in allusion to the color marking 
of the back. 

PARAPERCIS CLATHRATA Offilby. Plate 3, fi». 1. 

Percis Utracanthus LacepIjde, Hist. Nat. Poiss. 4*. (1803) 285, 302. 
Percis tetracanthus Bleeker, Nat. Tijdschr. Ned. Ind. 4 (1853) 458; 

GtJNTHER, Cat. Fish. Brit. Mus. 2 (1860) 241. 
Parapercis clathrata Ogilby, Proc. Roy. Soc. Queensl. 23 (1911) 41. 
Parcis quadrispinosus Weber, Siboga Exp., Fische (1913) 519. 
Parapercis tetracanthus Fowler, Mem. P. Bishop Mus. 10 (1928) 


Dorsal IV, 21; anal 18; scales on lateral line 59 or 60, be- 
tween lateral line and origin of dorsal 7; between lateral line 
and origin of anal 15 to 16. 

Body elongate, subcylindrical, its depth 5.7 to 6.5 in length; 
caudal peduncle compressed, its least depth 3.1 to 3.7 in head; 
dorsal profile more arched than ventral ; head slightly depressed, 
3.2 to 3.4 in length of body ; snout moderate, gently sloping to 
tip, 2.8 to 3 in head; eye moderate, gently sloping to tip, 2.8 to 3 
in head ; eye moderate, nearer tip of snout than angle of oper- 
cular opening, its upper margin projecting above outline of 
head, 4 to 4.3 in head ; interorbital narrow, 10.8 to 16 ; mouth 
moderate, oblique, lips slightly broad, lower jaw projecting; pre- 
maxillaries protractile ; maxillaries entirely concealed under pre- 
orbital, shorter than snout, and sometimes reaching a vertical 
through anterior border of eyes; upper jaw with an outer series 
of coarse teeth with several long ones in front and sides, and 
with an inner villiform band narrowing posteriorly; lower jaw 
with an outer series of coarse teeth, four in front and three at 
each side, all enlarged and directed posteriorly, and an inner 
villiform band which is not continuous behind; vomer with a 
small patch of small teeth; preopercle entire, opercle with a 
strong rather flat spine at posterior angle. 

Dorsal continuous, third spine longest and as high as vertical 
diameter of eye; longest rays of dorsal as long as snout or a 
little longer; anal inserted below fifth dorsal ray; pectoral 

5«,2 Martin and Montalban: Philippine Parapercidx 223 

rounded, 4.6 to 5.5 in length; ventrals pointed, 4 to 4.57 in 
length of body. 

Scales behind occiput, on cheeks, breast, and abdomen smooth, 
those on body and caudal finely ctenoid ; scales absent on snout, 
interorbital, sides of head, and occiput. 

The following color notes were taken from a live specimen, 
107 millimeters long, from Basco, Batanes Province. Ground 
color brown, each side of body with ten vertical bars which are 
brown above and red below; vertical bars crossed by two 
longitudinal bars, upper brown and lower red with black spots; 
interspaces between vertical and longitudinal bars yellowish, 
lower parts of body fading into white. Top of head with large 
round black spots; sides with much smaller ones. Upper part 
of body brownish with rather indistinct vertical crossbands. 
Caudal with few black short bars and spots, and brown vertical 
bars. Dorsal colorless, with two rows of black spots ; anal color- 
less, with black spots behind bases of 6th, 9th, 12th, and 15th 

Color in alcohol yellowish to dusky brown, becoming pale 
brown to white below ; top of head with blackish spots, a black 
ocellus with narrow border of white often appearing only as a 
black spot immediately above angle of opercular opening; a 
brown to dusky brown blotch on cheeks below eyes; posteriorly 
a lighter blotch with black dots, faded in three specimens ex- 
cept for the dots; eleven rather faint crossbands on back, the 
first connecting the two ocelli on suprascapula and the last on 
the caudal peduncle; on each side nine vertical bars not con- 
tinuous with the crossbands on the back; three faint brownish 
longitudinal bands on each side, one running from the angle of 
opercle to upper third of caudal peduncle, connected with the 
crossbands on back and the vertical bands on sides, the second, 
usually fainter, running straight from upper part of axilla to 
midst of caudal peduncle, and the third one, absent in some spec- 
imens, from immediately below root of pectoral straight to 
lower part of caudal peduncle; this band is narrower and con- 
nects a series of nine to ten black spots on lower part of the 
vertical bars the first of which lies below root of pectoral and 
the last on lower part of caudal peduncle; pectoral and ventral 
white; dorsal with three series of black spots at base of anal 

224 The Philippine Journal of Science 1935 

which are absent or very faint in some specimens, and a series 
of smaller faint spots near the outer edge; caudal with a small 
white blotch. 

Here described from four specimens, 99 to 152 millimeters, 
from the following localities : Bataan, Batanes Province, Basco. 
Camiguin, Cagayan Province, Camiguin. Leyte, Leyte Prov- 
ince, Cabalian. Balabac, Palawan Province, Balabac. 

We adopt the nom. subst. of Ogilby by priority over the nom, 
nov. of Weber in accordance with Opinion 1 rendered by the 
International Commission of Zoological Nomenclature. 

PARAPERCIS MONTILLAI sp. nov. Plate 3, fig. 2. 

Dorsal IV, 21 ; anal 18 ; scales in lateral line 60, between lat- 
eral line and origin of dorsal 7, between lateral line and origin 
of ventral 14 to 16. 

Body elongate, rather cylindrical, its depth 6 to 6.6 in length; 
caudal peduncle compressed, its least depth 3.2 to 3.5 in length 
of head which is 3.2 to 3.4 in length of body; snout rounded at 
tip, 3 to 3.6 ; eyes moderate, nearer to tip of snout than to hind 
edge of opercle, 4.5 to 5 ; interorbital space very narrow, 17.5 to 
22; mouth moderate, oblique; lips slightly broad; premaxillaries 
protractile; maxillary concealed by preorbital, extending below 
anterior border of eye; lower jaw projecting; upper jaw with 
an outer series of enlarged teeth, which are coarse and canine- 
like anteriorly, and with a band of villiform teeth; lower jaw 
with three curved canines on each side of symphysis and a 
villiform band, which becomes a series of coarse teeth poste- 
riorly; vomer with a narrow villiform band; a spine at upper 
angle of preopercle; subopercle terminates in a rather flat, 
pointed spine or in a sharp edge. 

Dorsal continuous, third spine longest, as long as eye diam- 
eter; anal inserted below fifth dorsal ray; ventral pointed and 
short, 4 to 4.75 in length of body ; caudal much rounded. 

Scales ctenoid; smooth behind occiput, on cheeks, breast, and 
belly; snout, and top and sides of head, naked; basal half of 
caudal scaly. 

This species differs from P. clathrata in its more rugged ap- 
pearance and in the more defined outline of the markings, which 
run from dusky brown to brownish and are arranged more or 
less in the form of a series of chevrons. Suprascapular ocelli 
are absent. The markings on the head are in the form of brown 
blotches. A faint blotch on the anterior portion of the cheek 
and a darker one behind with three oblique white lines. The 

56, 2 Martin and Montalban: Philippine Parapercidse 225 

black blotch at lower base of the pectoral forms a line with a 
series of seven to eight others on side of the belly to the lower 
side of the caudal peduncle. A posteromedian white blotch on 
the caudal. 

Type: No. 31284, 137 millimeters long, in the Bureau of 
Science collection, collected by Jose Montilla, from Calapan, 
Mindoro, 1931. Four cotypes: One specimen, No. 4382, 134 
millimeters long, collected from Zamboanga, Mindanao, June 
10, 1908, by Alvin Seale; three others. No. 11438, 90 to 154 
millimeters long, collected from Calapan, Mindoro, by Gregorio 
Lopez, January 17, 1923. 

Montillai, for Jose Montilla, ichthyologist, Bureau of Science, 

289858 9 


Plate 1 

Fig. 1. Parapercis cylindrica (Bloch). 

2. Parapercis tetracantha (Lacepede), 

Plate 2 

Fig. 1. Parapercis hexophthalma (Cuvier and Valenciennes )« 
2. Parapercis dorsonebulosa sp. nov. 

Plate 3 

Fig. 1. Parapercis clathrata Ogilby. 
2. Parapercis montillai sp. nov. 




The Philippine 
Journal of Science 

Vol. 56 MARCH, 1935 No. 3 


By Joaquin Maranon 
Of the Bureau of Science, Manila 

Amando Perez 

Of the Bureau of Prisons, Manila 


Paul F. Russell 

Of the International Health Division of the Rockefeller Foundation 


Totaquina is the name suggested in 1931 by the subcommittee 
of experts of the Health Organization of the League of Nations (i) 
for a new standard preparation of the total alkaloids of cinchona 
bark. This product is recommended as a cheap remedy for 
malaria in places where the cost of quinine is prohibitive. Our 
paper reports some clinical results following the use of tota- 
quina extracted at the Bureau of Science from cinchona bark 
of trees grown by the Bureau of Forestry in Bukidnon, Min- 
danao. In other words, the drug discussed in this paper is 

^ These studies were carried out with the cooperation of the Bureaus of 
Science, Prisons, and Forestry, together with Malaria Investigations, which 
is jointly supported by the Bureau of Science and the International Health 
Division of the Rockefeller Foundation. The cinchona bark was supplied 
by the Bureau of Forestry through the courtesy of Director A. F. Fischer. 
Totaquina was extracted at the Bureau of Science and was tested clini- 
cally at the Iwahig Penal Colony of the Bureau of Prisons, with the assist- 
ance of the staff of Malaria Investigations. 

The writers desire to express their thanks to Dr. Mariano V. del Ro- 
sario, director of the School of Pharmacy, University of the Philippines, 
for the use of a drug mill in powdering the cinchona bark. Also to For- 
ester Agapito L. Cenabre for information regarding cinchona bark in the 

290499 229 


The Philippine Journal af Science 


entirely a Philippine product, produced in experimental quan- 
tities. This paper, therefore, suggests a new and potentially 
important local industry which might have far-reaching effects 
in combating malaria in the Islands. 

One of us (2) has estimated that there are probably at least 
2,000,000' cases of malaria a year in the Philippines. Assuming 
that the average quinine requirement per case is 250 grains 
(a low figure), it was stated that the Islands could theoretically 
use about 32,400 kilograms. Actually, the annual importation 
of quinine is frequently less than 2,000 kilograms (Table 1). 

Table 1. — Quantity of quinine imported into the Philippine Islands, 



Amount. ' 




Amoun « . 



b 1,489 

b 10,515 

b 8,227 

b 7,226 


b 7,724 

b 11,857 

b 12,865 

b 10,841 
















1923 _ 







d 28,375 

d 21,442 

d 23,410 

d 20,662 

d 14,839 















1928 _-_ 

1905 __ 






1907 _ _ 



1908 _ 


1932 - 





*This table is taken from Russell. (2) The figures, supplied by the Bureau of Cus- 
toms, are the best available; but it must be noted that cinchona bark and quinine in 
various forms are included in the total. Several thousand hectograms of bark are said 
to be imported each year so that the figures given above are greater than the actual amount 
of the alkaloid quinine actually imported. The price of quinine sulphate has varied greatly 
and there are no records available in government or private files. The present wholesale 
price of quinine sulphate is 43 pesos (21.50 dollars United States currency) per kilogram 
(June, 1934). It is imported direct from Java. 

^ Estimated on the basis of value. Customs reports give only value and not amounts 
for these years. 

<= Importations of quinine not separately recorded for these years. 

d Courtesy of the Bureau of Customs. 

The estimate of 250 grains per case per year is lower than 
that of the League of Nations Health Organization, (3) which 
used 20 grams or about 308 grains per year as an average 
figure. But the league report estimates the quinine require- 
ments of the Islands at only 1,305 kilograms as against the 
estimate of 32,400 kilograms quoted above. This discrepancy is 
due to the fact that in the league report the number of cases 
of malaria in the Philippines in 1930 is given as 64,251, an 

56»3 Maranon et ah: Totaquina 231 

impossible figure in view of the reported number of 15,144 
malaria deaths in the same year. The disease malaria does not 
kill directly anything like 25 per cent of its victims. The deaths 
are probably not more than 1 or 2 in every 200 eases. There- 
fore, to account for 15,144 deaths there must have been from 
1,500,000 to 3,000,000 cases, assuming the diagnosis of cause of 
death to have been accurate (which, of course, is often impos- 
sible). This question is discussed by one of us (2) elsewhere. 
We do not claim to know the correct figure, but we believe that 
at least 32,400 kilograms of quinine would be required to treat 
properly the cases of malaria now occurring each year in the 
Philippines. If this be true and if present quinine importations 
are continued, there remains a potential local market for at 
least 30,000 kilograms, more than 33 short tons, per year of 
locally produced totaquina. 

It is a matter of common observation that very few cases of 
malaria in the provinces are adequately treated. The price of 
quinine sulphate varies but is commonly 1 centavo (0.5 cent 
United States currency) or more per grain as sold at retail, 
although it sells in bulk wholesale for 43 pesos (21.50 dollars) 
per kilogram. The retail price of one 5-grain tablet of quinine 
dihydrochloride is as high as 15 centavos in provincial boticas. 
At the usual retail prices a 250-grain treatment with quinine is 
out of the question for many thousands of malarious persons 
in the Philippines. Refined quinine is, in fact, an expensive 
luxury, a rich man's remedy. So too is atabrine, which retails 
for about 2.50 pesos per 15-tablet treatment. Totaquina might 
meet the need for a poor man's febrifuge, especially if it could 
be produced locally. 


It would be interesting to recount at length the history of 
cinchona bark, but in this paper it must suffice merely to outline 
some of the important facts. This subject is admirably pre- 
sented by Suppan.(4) According to this historian there is no 
solid foundation to support the interesting legend that we owe 
our first knowledge of the febrifuge qualities of cinchona to 
Peruvian Indians. For instance, from the discovery of Peru 
by Spaniards in 1513 to the beginning of the seventeenth 
century, none of the observers who wrote of their experiences 
in South America mentioned in any way a febrifuge bark. More- 
over, the Indians themselves, after the Spaniards had learned 
of this bark, refused to use it as a medicine even after the 
properties were explained to them. 

232 The Philippine Journal of Science inn 

According to Suppan, it is highly probable that Jesuit mis- 
sionaries were the first to discover the virtues of cinchona bark 
at some date and place still unrevealed. The first-known use 
of powdered cinchona bark to cure malaria was in 1630 when 
Don Juan Lopes de Cannizares, Spanish corregidor of Loxa, 
Peru, was successfully treated. In 1638, this corregidor sent 
some bark for the treatment of Countess Chinchon, wife of the 
viceroy of Peru. Having first tried it on a number of patients, 
her physician, Dr. Juan de Vega, gave some to the countess, 
with brilliant results. She was so grateful that she secured 
another supply and gave it gratis to all who applied at her 
palace. Hence its first popular name, polvo de la Condesa, So 
many applied for this remedy that she turned over to Jesuits 
the task of distributing it. The name then became polvo de los 
Jesuitos or polvo de los padres. 

It is not clear exactly when the bark was first sent to Europe, 
but soon after 1640 it was widely known, largely because of the 
activity of Jesuit fathers, among whom Cardinal de Lugo was 
most enthusiastic. Hence another name — pulvis cardinalis. In 
1649 this cardinal cured in Paris the dauphin who later became 
Louis XIV. This success naturally called widespread attention 
to powdered bark, which was becoming commonly known as 
pulvis Peruanus or Peruvianum fehrifugum. 

Just when the powdered bark first came to the Orient is 
uncertain, but it is recorded that in 1692 Jesuit priests in 
China used it to cure Emperor K'ang Hsi of a dangerous fever. 
It is not unlikely that supplies of the bark had also come to the 
Philippines at this early date. The Jesuits were absent from 
the Philippines from 1768 to 1852, but it seems very probable 
that they had imported supplies of this febrifuge prior to their 
departure. However, we have not been able to find any definite 
record of this. The fact that Filipinos have used bitter pow- 
dered bark from such local trees as dita \_Alstonia scholaris 
(L.) R. Br.] as a febrifuge for more than one hundred years, 
is probably accounted for by a similarity in taste to powdered 
Peruvian bark. 

Not until 1740 was a description of the fever-bark tree pub- 
lished in Europe. La Candomine, a French scholar, sent draw- 
ings and a description to Paris, and two years later Linnaeus 
established the genus Cinchona, He misspelled the name of the 
countess whom he sought to honor, but the International Bota- 
nical Congress held in Paris in 1866 voted to retain his spelling. 

56,3 Maranon et al.: Totaquina 233 

In 1752 the Spanish Government began to organize the bark 
trade. It also searched for cinchona trees in regions other than 
Peru, finding them in Chile, Ecuador, and Bolivia. It may be 
noted that cinchona trees have never been found growing natu- 
rally in any part of the world other than this limited South 
American region. 

The researches done by Humboldt and by Weddell were of 
great importance in spreading information regarding the col- 
lection and preparation of the barks and their botanical classi- 

In 1820 two French pharmacists isolated quinine and cincho- 
nine from the bark. The discovery of quinine, in particular, 
led to an enormous increase in the demand for bark. There fol- 
lowed a rapid and reckless destruction of cinchona trees, so that 
the supply was in danger of exhaustion. This suggested the 
possibility of cultivating cinchona elsewhere. Java was sug- 
gested in 1837 and India in 1839. 

The Dutch were first to bring cinchona trees to the Orient. 
Hasskarl, superintendent of gardens in Java, left Peru in 1854, 
arriving in Batavia after a journey lasting four months. All 
but two of his plants died during the journey or afterwards. 
However, with his surviving specimens, augmented by later con- 
cessions, he was successful, and was knighted for his achieve- 

The British had tried unsuccessfully in 1852 to raise cinchona 
from seeds sent to Calcutta; but in 1860, due to the eiforts of 
Markham and Spruce, they were able to start growing cinchona 
in India. Their supplies came partly from South America and 
partly from Java. By 1866 there were more than 1,500,000 cin- 
chona plants growing on the Nilgiri Hills, in southwest India, 
and this number had nearly doubled by 1872. In 1861 another 
cinchona plantation had been started in Ceylon. 

The cultivation of cinchona in Java and India has continued, 
Java producing about nine-tenths of the bark of the world. 
The chief species now of commercial importance are Cinchona 
ledgeriana Moens, C. calisaya Weddell, C. officinalis Linnseus, C. 
hybrida, and C. succirubra Pavon. The first named has the 
highest quinine content. 

About thirty alkaloids have been isolated from cinchona bark. 
The four chief alkaloids have been isolated as follows : 

Quinine, 1820, isolated by Pelletier and Caventou. 
Cinchonine, 1820, isolated by Pelletier and Caventou. (Gomez, 1810, 
made a partial separation of cinchonine.) 

234 The Philippine Journal of Science 1935 

Quinidine, 1846, isolated by van Heijningen. 
Cinchonidine, 1847, isolated by Winkler. 

In both Ceylon and India the cultivation of cinchona was 
largely given up in order to produce the more-profitable tea, so 
that at present there is in reality a Dutch monopoly. Lately 
there has been a revival of interest in cinchona planting in 
India. Experimental cinchona is grown in certain areas of 
Africa, Indo-China, Malaya, and the Philippines, and in a few 
other regions. It does not appear that any unfair advantage 
has been taken of their monopoly by the Dutch. In fact there 
has been complaint about overproduction with insufficient returns 
on the investment. Certainly the Dutch have kept pace with 
the demand for quinine and have made every effort to reduce 
the costs of production. 

Nevertheless, there is a serious world shortage in terms of 
the amount required by the estimated number of malaria pa- 
tients in the world to-day. This is due in part to faulty dis- 
tribution in malarious regions, but the chief reason is the fact 
that millions of sufferers cannot afford to buy quinine even at 
its relatively low price. Therefore, there is a very large field 
for the production of a still lower-priced product, such as tota- 
quina. As noted above, there is a potential market in the 
Philippines alone for more than 30,000 kilograms a year. Tota- 
quina cannot be considered to be a direct competitor of either 
quinine or the synthetic antimalaria drugs. It is required by 
malarious consumers who literally cannot afford to buy higher- 
priced drugs. 


The first reference we have found to cinchona growing in 
the Philippines is contained in a short note (22) about the ''Jar- 
din Botanico." The director of this garden in 1893 is reported 
to have said that he would like to acclimatize "Sinconas" at the 
foot of the Banahao mountains in Tayabas and also in Munang 
near Antipolo. There is nothing in this note to indicate that 
this was done, and we have found no further reference to it. 

According to Fischer and Cenabre(5) it is safe to state that 
cinchona can be grown in the Philippines on areas properly 
located. They note that although it is commonly believed that 
cinchona is very delicate to handle, according to their ex- 


Maranon et aL; Totaquina 


perience and careful observation, the problem is not exceptionally 

It appears that the first attempt to grow cinchona in the 
Islands was made in 1912 in Baguio, Mountain Province, 
and in Los Banos, Laguna, by the Bureau of Forestry. These 
attempts failed, as did several later ones. In 1916 three 
hundred plants were set out, but very few survived. In this 
same year Wester, (6) a horticulturist, stated, after visiting some 
cinchona plantations in Java, that the soil and climate of some 
parts of southeastern Bontoc appeared to be favorable for cin- 
chona. He suggested that the Government consider the intro- 
duction of cinchona in those areas. It does not appear that his 
suggestion was followed, but an anonymous note (7) of 1919 
stated that the introduction of cinchona plants into the Islands 
from India was being attempted by "the Igorot Exchange, a 
missionary institution in Sagada." Some 10 ounces of seed of 
C. ledgeriana was received from, Ootacamund, Madras. Accord- 
ing to the Right Reverend Bishop Mosher, this experiment ap- 
parently did not succeed, as no evidence of it remains in Sagada. 

In 1927, according to Fischer and Cenabre(5) and Altami- 
rano,(8) the Bureau of Forestry, with funds made available by 
the Reforestation Act, started a cinchona nursery in Impalutao, 
Bukidnon, Mindanao, using seeds from Java. This area has an 
altitude of about 762 meters, with an average rainfall of 112 
inches annually, and temperature ranging from 17° to 29° C. 
The attempt has been successful, and an area of nearly 14 hec- 
tares has been planted to cinchona. There were 67,000 seedlings 
in the nursery ready for transplanting in 1933. There are now 
about 38,000 trees on the plantation itself, some 11,000 bearing 
fruit since 1930, and some 1,100 three or more years old. The 
species present are C. ledgeriana, C. succirubra, and C. hybrida. 

Some bark from 5-year-old trees was analyzed with the fol- 
lowing comparative results: (9) 

Table 2.-^inchona hark analyses. 


Percentage of alkaloids. 




Cinchona ledgeriana.. 
Cinchona hybrida - . 

Per cent. 

Per cent, 

Per cent, 

Cinchona tuccirubra. . 



236 The Philippine Journal of Science 1935 

A second area in Bukidnon has been planted. This is at 
Barrio Alanib, in Malaybalay, and has an elevation of 1,067 to 
1,372 meters. The trees are growing even better in this second 

Quoting former Vice-Governor-General C. Butte, (9) who visit- 
ed the plantations : "I can conceive of no production that would 
be more beneficial to all the people in these Islands, both from 
the commercial and the public standpoint, than the production 
of quinine within our country." 

The authors of this paper would subscribe to this statement 
provided for "quinine" is read "totaquina." It is doubtful if 
the Philippines or any other country could successfully compete 
with the Dutch in the manufacture of the separated alkaloid 
quinine, but totaquina could certainly be produced and would 
find a very large potential market within the Islands. 


As a result of several conferences and considerable research 
the Malaria Commission of the League of Nations (lO) suggested 
that a new cinchona product be prepared. This is to be called 
"totaquina," and is to be a standardized mixture of the com- 
bined alkaloids in cinchona bark, as follows : 

CrystalUzable alkaloids, — Not less than 70 per cent. Not less than 15 

per cent of these alkaloids to be quinine. 
Amorphous alkaloids. — ^Not to exceed 20 per cent. 
Moisture, — Not more than 5 per cent. 
Mineral matter, — ^Not more than 5 per cent. 

Two types of totaquina have been prepared in commercial 
practice, as follows: 

Type L — Made by extracting and precipitating as an almost 
white powder the total alkaloids from the bark of C. succirubra 
or C. robusta, which can be cultivated in many malarious coun- 
tries. This is the preferred type. It is the one we have in 
mind for the Philippines. 

Type 11. — Utilizing the residue remaining after quinine sul- 
phate has been extracted from C. ledgeriana and bringing the 
preparation up to the required standard by adding suflScient 
quinine and other crystallizable alkaloids. 

In this connection Howard, (ii) who was in a position to 
know, stated that some of the advocates of cheap febrifuge 
mixtures appear to believe that there are vast quantities of 
cinchona products being wasted, which could be used (as in 
Type II totaquina). Howard said that he had no knowledge 


Maranon et aL: Totaquina 


of the existence of any hoards of residual alkaloids anywhere 
in the world. 

The commission pointed out that the term **cinchona febri- 
fuge," which was as heretofore applied to various mixtures of the 
alkaloids, should be discarded in favor of standardized totaquina. 
The name "quinetum" should be used only for a preparation con- 
taining quinine, cinchonidine, and cinchonine in equal parts. 

The commission (10) reports that Giemsa(i2) found totaquina 
a little more toxic than hydrochlorate of quinine when given in 
the same doses intravenously to rabbits. He found totaquina, 
both types, somewhat inferior to quinine in avian malaria (Plas- 
modium relictum) . Giemsa, after a number of tests, concluded 
that the therapeutic efficacy of totaquina was directly propor- 
tional to the amount of quinine present. 

Also cited by the commission are some tests by James (10, 13) 
in cases of benign tertian malaria intentionally induced by the 
bites of mosquitoes. The samples he used were composed as 
follows : 

Table 3. — Composition of samples used in tests by James. 


Type I. " 

Type II. b 

Quinine _ 

Per cent. 





Per cent. 





Cinchonine _ . _ 

Cinchonidine _. _ . 


Total crystalllzable alkaloids 






Amorphous alkaloids __ _. _ _ . 



* Type I consisted of total alkaloids of C. auccirubra. 
^ Type II was composed of auinine residues. 

James's(iO) results may be quoted as follows: (a) One dose 
of 5 grains (0.3 gram) of totaquina of either type has practi- 
cally no effect on the fever or parasites. It is necessary, there- 
fore, to use a single dose of 10 grains (0.6 gram) for the test. 
(&) A single dose of 10 grains of totaquina Type I produces 
the same effect in aborting the fever and in reducing the para- 
sites as that produced by a single dose of 5 grains of quinine, 
(c) A single dose of 10 grains of totaquina Type II has prac- 
tically no effect in aborting the fever or in reducing the para- 
sites. With this type of totaquina it is necessary to use a single 

238 The Philippine Journal of Science 1935 

dose of 20 grains (1.2 grams) to produce the same effect as that 
produced by a single dose of 5 grains of quinine. 
James (10) concludes as follows: 

It appears from these results that totaquina Type I (total alkaloids of 
C. succirubra or C robusta) when used in ordinary clinical therapeutic 
doses for curative purposes — e. g. 1.2 gram (20 grains) daily for five to 
seven days — should give about the same good result as is given by qui- 
nine in the same doses. If this is so, and if this type of totaquina can 
be obtained more cheaply than quinine, it would be advantageous to use 
it for general purposes instead of the single alkaloid. 

The method by which Type I in the above tests was prepared 
is given by Groothoff and Henry (13) as follows: 

The process used was that described by Schwyzer, Fabrikation der Al- 
kaloide, Berlin (1927), which depends on the extraction of the alkaloids 
from the ground bark, previously mixed intimately with slaked lime, by 
boiling benzene. From the benzene solution the alkaloids are in time 
removed by agitation with dilute sulphuric acid. The aqueous alkaloidal 
solution so obtained is decolourized by heating with charcoal, filtered, and, 
while still hot, run into excess of sodium hydroxide solution to precipitate 
the alkaloids. After standing for several hours the precipitate is collected, 
washed with water and allowed to dry in the air. 

The samples obtained were nearly white or pale cream-colored 
powders, completely soluble in dilute sulphuric acid, giving a 
fluorescent solution. 

The British Pharmacopoeia (14) includes totaquina (tota- 
quine), giving its characters, tests for identity and purity, and 
its assay. 

According to Dawson, (15) who cites Henry, (16) the four 
common cinchona alkaloids are really two pairs of optical iso- 
mers, quinine and quinidine forming one pair, and cinchonidine 
and cinchonine the other. Quinine and cinchonidine are levoro- 
tatory, and quinidine and cinchonine are, respectively, their 
dextrorotatory isomers. Dawson, (15) after reviewing the liter- 
ature, states that quinine and quinidine are of practically equal 
value in the therapy of malaria of any species. All four alka- 
loids are about equal in their efficacy against chronic benign 
tertian malaria. 

Totaquina would replace such preparations as quinetum, first 
extracted in Java in 1874, which was not standardized and did 
not prove to be economical. It was a coarse preparation of 
mixed alkaloids. So, too, a cinchona febrifuge similar to but 
better than quinetum was never well standardized. The Indian 
cinchona febrifuge came to be a residual alkaloidal rather than 

56.3 Maranon et aL; Totaquina 239 

a total alkaloidal preparation. Nevertheless, cinchona febrifuge 
has had considerable use and value. It has become the most 
important of the commercial alkaloid preparations. 

^ It may be noted that the usefulness of the other alkaloids of 
cinchona in malaria therapy has been well known since 1866- 
1867 when three Madras Commissions in India reported favor- 
ably after extensive experiments. There has been ample con- 
firmation by Fletcher (17) and others. Field (18) has recently 
reported some observations that led him to conclude that tota- 
quina gives excellent results in malaria therapy. The subject 
is also reviewed by Sinton,(l9) especially as to the place of 
cinchona febrifuge in malaria therapy. 


The totaquina used in this investigation was prepared from 
the ground bark of Cinchona ledgeriana, which was grown at 
the Philippine Bureau of Forestry Station at Impalutao, Bukid- 
non, Mindanao. As previously stated, the bark of this Cinchona 
species has a higher percentage of total alkaloids than those 
of C. succirubra and C. hybrida, which are also grown there. 

The results of our analysis of Philippine Cinchon/i ledgeriana 
bark, according to the method of Howard and Chick, (20) are 
given! in Table 4. As shown by the data, this bark has a rather 
high percentage of total alkaloids, more than half of which is 

Table 4. — Analysis of the dried bark 

of Philippine 

Cinchona ledgeriana. 


Per cent. 









Amorphous alkaloids 


Total alkaloids 9.62 

Preliminary experiments were carried out in order to ascer- 
tain a satisfactory method for making totaquina from cinchona 
bark. As a result of our experiments, we made three prepara- 
tions which are recorded as A, B, and C. 

The method used for making totaquina A was essentially the 
same as that recommended for the preparation of "quinetum" 
from Cinchona succirubra A^^) The result was not so satis- 

240 The Philippine Journal of Science isss 

factory, as we obtained a dark brown powder that was only 
partly soluble in dilute sulphuric acid. 

Our preparation of totaquina B gave better results. The 
method for this preparation was very similar to that employed 
in making totaquina A, except that the precipitated alkaloids 
were dissolved in dilute sulphuric acid (about 0.2 N) and fil- 
tered to remove insoluble impurities. The acid filtrate was 
heated and then decolorized with purified charcoal. The hot 
solution was then made strongly alkaline with sodium hydrox- 
ide solution in order to precipitate the alkaloids. When cooled, 
the precipitate was collected on a filter, washed with water, and 
dried in a vacuum oven below 50° C. Totaquina B has a light 
brown color but is not completely soluble in dilute sulphuric acid, 
though much more nearly so than totaquina A. 

Our third preparation, totaquina C, gave the best results. 
This product was made in the following manner: The finely 
powdered cinchona bark was macerated with dilute hydro- 
chloric acid. The mixture was then poured into a percolator and 
the extraction completed with water as the solvent. The acid 
percolate was treated with a sufficient amount of sodium hydrox- 
ide solution to precipitate the coloring matter as flakes. After 
the coloring matter had been removed, the solution, which was 
still strongly acid, was made strongly alkaline with sodium hy- 
droxide solution. The mixture was set aside overnight and the 
alkaloids that separated from the turbid liquid were removed by 
filtering. They were washed with water until the washings 
were almost colorless, and dried. This represented the first por- 
tion of the totaquina. The filtrate from the first portion con- 
tained a considerable quantity of alkaloids in suspension. These 
were precipitated (salted out) with sodium chloride. After 
standing for several hours, the supernatant liquid was removed 
and the precipitate collected on a double filter and washed with 
water. The precipitate was treated with dilute sulphuric acid 
and filtered to remove impurities. The acid filtrate was de- 
colorized with animal charcoal and the hot solution precipitated 
with an excess of sodium hydroxide. The precipitate was re- 
moved, washed with water, and dried, after which it was ex- 
tracted with hot ethyl alcohol (95 per cent). The alcoholic 
extract was distilled to remove the excess alcohol. The residue 
was evaporated to dryness and incorporated with the first frac- 
tion of the alkaloids previously obtained. A kilogram of pow- 
dered cinchona bark gave 90 grams of totaquina C or a yield of 


Maranon et al: Totaquina 


9 per cent. A complete assay of the bark (Table 4) gave 9.62 
per cent of total alkaloids. The yield of totaquina C is, there- 
fore, about 93 per cent of the total alkaloids in the bark. 

This product, totaquina C, was a pale brown powder, which 
dissolved easily and completely in dilute sulphuric acid, giving a 
blue fluorescent solution. 

The alkaloidal constituents of totaquina B and C were deter- 
mined according to the method of Howard and Chick. (20) The 
results are recorded in Table 5. Instead of determining the 
quinine by weighing the dried quinine sulphate, we dissolved the 
salt in hot neutral ethyl alcohol and titrated the solution with 
alcoholic potassium hydroxide (0.1 iV), using phenolphthalein 
as indicator. 

Table 5. — Analysis of totaquina preparations from the bark of Philippine 

Cinchona ledgeriana. 






Per cent. 

Per cent. 

Cinchonidine __ 


Cinchonine ,_ 

Total crystallizable alkaloids . _ _ . _ 



Amorphous alkaloids _ 

Total alkaloids . .. 



Moisture __ 


Organic impurities (by difference) __. __. _ 

As shown by the data (Table 5) totaquina C is a better prod- 
uct than totaquina B because it contains not only more of the 
total crystallizable alkaloids but also more quinine. 

The cost of preparing totaquina C in our laboratory was ap- 
proximately 16.07 pesos per kilogram. This includes the aver- 
age cost of cinchona bark (87 centavos per kilogram), as given 
by the Bureau of Forestry (see below), and incidental expenses 
for chemicals and labor. The estimated expenses are itemized 
as follows: 

Cinchona bark, 11.1 kilos 9.66 

Concentrated hydrochloric acid, 610 cc 1.60 

Sodium hydroxide, 597 g 1.36 

242 The Philippine Journal of Science 1935 


Sodium chloride, 1 kg 0.20 

Concentrated sulphuric acid, 100 cc 0.20 

Charcoal, 500 g 0.65 

Ethyl alcohol, 1 kilo 0.20 

Filter paper, 10 sheets 0.20 

Total cost of materials 14.07 

Labor 2.00 

Total expenses 16.07 

This is a laboratory cost. In commercial production the cost 
would doubtless be less. As stated above the wholesale price of 
quinine sulphate in bulk is 43 pesos per kilo (the Government 
pays 40 pesos). In 5-grain tablets the wholesale price is 14 
pesos per 1,000 tablets (the Government pays 12 to 13 pesos) . 
This price for tablets is practically equivalent to the price for 
the powder. 

Boticas, or drug stores, in the provinces, charge 1 centavo a 
grain for quinine sulphate, as an average minimum. The price 
is not infrequently doubled in places where the demand is great- 
est. At 1 centavo a grain the retail cost of quinine sulphate 
becomes 154 pesos per kilogram. 

Obviously, totaquina can be sold for much less than that. We 
submit the following estimate of the retail cost of a kilogram of 
totaquina as a fair approximation : 


Cinchona bark 9.66 

(This allows a fair profit to the grower. See table 
Preparing the totaquina 7.37 

(This allows a 15 per cent profit on production 
costs of 6.41 pesos as set forth above.) 

Wholesale price per kilo 17.03 

Allowance for retailer's profit and expenses, 20 per cent 3.40 

Retail price of totaquina to consumer, per kilo 20.43 

This estimated retail price of 20.43 pesos per kilogram of tota- 
quina is higher than it would be if production were carried on 
commercially, but it compares very favorably with the present 
retail price of 154 pesos for quinine sulphate. If retailers were 
now limited to a 20 per cent profit, the retail price of quinine 
sulphate would be 51.60 pesos per kilogram, more than two and 
a half times the liberally estimated totaquina price. But since 
established prices are not easily changed it is perhaps fair to 

Day Day Day Day 
123456789 123456789 123456789 123456789 

11 -CI "C^ "C? . "^^ 

41. - - T * ■ fs^ * 

no J * . \ P 

:;- Ah ■ V/^ ■ ^\ ^, -^^^^^^icx 

37: -^r\:i:.-v^ y ^v*^ . [ v*^-o^v-^ ju^'^^Vv"'*^-- 

.,-C5 -C6 C? ^ -C8 

. 41 — T r- _ r— T- 

dOqc' /> - ^-i^ ^A \ 

m - „.-. -- ■, ■- - ' — 

§^,-08 -CIO^ -CU -CIZ^ 

if,- ^j^ ■ ^^^^ ■ ^^^r-«- ■ L^^ « 

Q 37 ]^pj----j--^— 5- . ^. - . .^^.,^J^^ — ^«^^X^Pq^, •— 1^^^ jp^ 

W ^ ^ ^ 

^^,-C13 -CI4 ^ -C15 , -C16 ^ 

^'"firt' - ^ " ^ " N ,f>, - --.. - \ A A ■■ --■ - 

§v- "^^^ ■ ^^ =^ • itV =. ■ "?^ta--- 

5 37 . V dB «^ w^ - _' -fc; 5B, s:*^ ^ . "^ vt' *** •'' ^ ^ w^ nits , 


C. , -r -, , , 

S_^,-C17 -C18 ^ C19 ^ CZO ^ 

S««" V lA " /V H " -Ar — n 

'^?^ ^^13 ■ J^itv^^ ■ -i2w„ =- ■ Aa^««« 

37^ M Ig-pj-j-jg:,,. *-^,i^^^ —iF— ¥1R3^^3W -^!5==*BS;*9^ 

r ?i 

41 f lOiacj^uiiici 1 noTJupy p _ Pn^^ 

2?- 5Ea^ Iwahig Penal Colony Hospital T = Rpst dose of totaquina. 

Oi' "■ >|^^/\,yAA^* lernpepcitupc i/hapts. 


Fig. 1. Temperature charts* totaquina therapy, Iwahig Penal Colony Hospital. Patients were hospitalized for at least fifteen days. The part of the 
290499. Facing page 242. temperature charts not shown above presented only normal temperature. 

56p3 Maranon et al: Totaquina 243 

contrast the present actual retail price of quinine of 154 pesos 
with the estimated price of 20.43 pesos for totaquina, which as a 
new drug being introduced by the Government could from the 
first be subject to some price restrictions under official supervi- 
sion. If this comparison is used the totaquina could be sold to 
the people for something like one-seventh of the present price of 
quinine ! In this case a 250-grain treatment would cost not 2.50 
pesos, as at present, but only about 35 centavos, a price within 
the means of the malarious poor. If sold to the people by the 
Government, eliminating the retailer's profit, the price would be 
even less. 


In this study of totaquina, twenty-one frank cases of malaria 
having positive blood examination for parasites were selected for 
trial. Particular attention was given to its effect upon chills and 
fever and upon schizonts and gametocytes of the two common 
types of malaria infections; namely, the benign tertian, caused 
by Plasmodium vivax, and the malignant tertian, caused by 
Plasmodium falciparum. Observations were made as to any 
untoward effects which could be attributed to the drug. Careful 
clinical histories as well as physical examinations of the cases 
for trial were noted, and blood examinations made at intervals 
before and during the administration of the drug. Thick and 
thin smears were stained with Giemsa, and each thin smear 
was examined for at least thirty minutes if negative. 


Totaquina was given in the same dosage as quinine in the treat- 
ment of malaria; namely, 0.6 gram three times a day for adults. 
It is not so bitter as quinine and hence can be administered to 
children in papers with lactose. Children less than 2 years of 
age were given 0.2 gram three times a day. Totaquina sample 
A was used to treat cases 1 and 2; sample B, cases 3 to 16, 19, 
and 20 ; sample C, cases 17, 18, and 21. 



Cases 1 to 18 were all diagnosed as benign tertian (case 13 was 
a mixed infection with subtertian). All were Filipino males 
unless otherwise stated. Fever charts are given in fig. 1, 
and a record of blood smear examinations in Tables 6 and 7, and 


The Philippine Journal of Science 


fig. 2. The physical examination was normal unless otherwise 


Clinical history. — Fever and chills accompanied by headache and profuse 
sweating regularly every afternoon for three days. Spleen three-fingers* 
breadth, palpable at left costal margin on normal respiration. 

Treatment. — ^Totaquina, 0.6 gram three times a day for fifteen days. 

Observation. — Two days after administration of the drug temperature 
became normal. Spleen receded in size until at the time of discharge it 
was palpable on deep inspiration. 


Clinical history. — Chills and fever every other day for three days. 
Spleen not palpable. 

Treatment. — ^Totaquina, 0.6 gram three times a day for fifteen days. 

Observation. — One day after administration of the drug temperature be- 
came normal. No more chills. The 3d day no malarial parasites could be 
seen in thick smear. 


Clinical history. — ^Fever and headache for two days. Spleen palpable on 
deep inspiration. 

Treatment. — Totaquina, 0.6 gram three times a day for ten days, then 
reduced to 0.3 gram. t. i. d. until the 15th day. 

Observation. — After administration of the drug there was no further 
rise of temperature. Two days after, the blood became negative for ma- 
larial parasites and remained so. The 12th day spleen no longer palpable. 


Clinical history. — ^Fever, chills, and headache every afternoon for four 
days. Spleen not palpable. 

Table 6. — Blood-smear examinations. Totaquina therapy. 

Colony Hospital." 

[-}-, Positive, but fewer than 1 per 100 leucocytes.] 

Iwahig Perwl 

Parasites per 100 








C. L 

Apr. 11 


Plastmodium vivax. 

Apr. 12 




..P 1 

Apr. 11 



Apr. 12 



R. S- J 

June 13 
June 14 




June 15 


c. c 

June 15 




June 16 


June 17 


a A complete record of negative examination is given in text fig. 2. 





















Totaquina Thepapy. 
I-vuahig Penal Colony Hospital. 







































s^ 1934 1 

<Juiiv 1934 










































































1 1 


1^1 1 






















1 1 1 1 1 1 1 1 i 











1 — Acs^^Yiinl 'nmon^i'fp'.^ 'Pmiindl 











- ® = Onlv aam^tocvtsrS found. 










- T =Fipst dose of totaauina.- 











P =FiPst dose, of Dlasmochin. 










1 I 

1 1 

1 1 

1 1 1 













































































































290499. FaciQiT paire S44. 

Fio. 2. Totaquina therapy, Iwahlir Penal Ckiionjr Hospital. Blood-smear chart. 


Maranon et al.: Totaquina 


Tabub 6. — Blood-smear examinations, Totaquina therapy, Iwahig Penal 
Colony Hospital," — Continued 










O. P. 

R. R.. 

F. C. 
A. P. 
F. G. 

F. A. 

P. C. 

S. M. 

B.d. V 

E ---- 




June 16 

June 16 

June 16 

June 17 

June 18 

June 16 

June 17 

June 18 

June 18 

June 19 

June 19 

June 20 

June 21 

June 19 

June 20 

June 21 

June 19 

June 20 

June 21 

June 22 

June 20 

June 21 

June 22 

June 25 

June 26 

June 27 

June 28 

July 8 

June 28 

June 29 

June 80 

July 1 

June 80 

July 1 

June 30 

•A complete 
290499 -2 



of negative 

Parasites per 100 










































Plazmodiwm vivax. 






Mixed Plasmodium vivax 
and /alctpafttw. 

PlawmodiMm vivax. 



examinations is driven in text fig. 2. 


The Philippine Journal of Science 


Tablb 6. — Blood-smear examinations, Totaquina therapy. Iwahig Penal 
Colony Hospital," — Continued 

Parasites per 100 







June 26 


Platmodium vivax and fal- 

June 27 



June 28 



June 29 



June 80 



July 1 



July 2 



July 8 



S. M 

July 4 
July 5 



July 6 


July 7 


July 8 


July 9 


July 10 


July 11 


July 12 


July 18 


July 14 


June 29 



June 80 


July 1 

+ . 

July 2 

July a 




July 4 
July 6 

July 10 


July 11 

July 12 


July 13 


July 14 


July 6 



July 7 



July 8 



July 9 




M. C 

July 10 
July 11 



July 12 


July 13 


July 14 


July 15 


•A complete record of negative examinations is given in text fig. 2. 

Treatment, — ^Totaquina, 0.6 gram three times a day for eight days, then 
reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — Two days after administration of the drug temperature 
became normal. The 4th day blood smear became negative. 


Maranon et aL: Totaquina 



Clinical history. — ^Fever, headache, and chills with profuse perspiration 
each afternoon for two days. Spleen palpable on deep inspiration. 

Treatment. — ^Totaquina, 0.6 gram three times a day for nine days, then 
reduced to 0.3 gram t. i. d. until the 15th day. 

Observation. — ^Two days after administration of the drug temperature 
became normal and blood negative. After ten days^ treatment spleen re- 
ceded and could no longer be palpated on deep inspiration. 


Clinical history. — Fever and headache, vomiting for two days. Spleen 
not palpable. 

Treatment. — ^Totaquina, 0.6 gram three times a day for seven days, then 
reduced to 0.3 gram until the 15th day. 

Observation. — Two days after administration of the drug fever subsided 
and blood became negative the next day. 

Table 7. — Disappearance of parasites following totaquina therapy." 

Parasites disappeared — 




In 1 day 





In 2 days _ 

In 3 days 

In 4 days _ 

In 5 days _ 

In 6 days.- 

In 8 days.- 

In 10 days 

In 18 days 




* In the history records the day is usually counted from that of admission. In the 
above table it is counted from the time the first dose of totaquina was given. 


Clinical history. — Fever and chills followed by profuse perspiration every 
afternoon for three days. Spleen palpable on deep inspiration. 

Treatment. — ^Totaquina, 0.6 gram three times a day for seven days and 
reduced to 0.3 gram t. i. d. until the 15th day. 

Observation. — Two days after administration of the drug temperature 
became normal and blood smear negative the following day. Spleen still 
palpable on deep inspiration, at time of discharge. 


Clinical history. — Headache, fever, and chilly sensation for two days. 
Spleen palpable on deep inspiration. 

Treatment. — Totaquina, 0.6 gram three times a day for seven days and 
reduced to 0.3 gram t. i. d. until the 15th day. 

248 The Philippine Journal of Science 


Observation, — The drug was administered at the time of subsidence of 
fever, which did not rise again. The next day blood smear became neg- 
ative. The 6th day spleen no longer palpable on deep inspiration. 


Clinical history,, — ^Fever, headache, and chills in the afternoon for two 
days. Spleen not palpable. 

Treatment, — Totaquina, 0.6 gram three times a day for seven days, and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — Two days after administration of the drug temperature 
became normal. Blood negative the 4th day. 


Clinical history. — ^Fever, headache, and chills coming every afternoon reg- 
ularly for three days. Spleen not palpable. 

Treatment, — ^Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — After three days' administration of the drug temperature 
went to normal, and blood smear became negative. 


Clinical history. — Fever, headache, and chills for two days each morning. 
Spleen not palpable. 

Treatment, — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — Three days after administration of the drug temperature 
became normal. Blood became negative the 5th day. 


Clinical history. — Fever, headache, and chills, followed by profuse sweat- 
ing coming every other day. Spleen palpable on inspiration. 

Treatments — Totaquina, 0.6 gram three times a day for seven days and 
reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — Three days after administration of the drug temperature 
became normal and blood smear negative. The 9th day, spleen could no 
longer be palpated on inspiration. 


Clinical history. — ^Fever, headache, and chills coming each afternoon for 
two days. Patient had chills in the ward. Spleen three fingers' breadth 
below left costal margin. 

Treatment, — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation. — Mixed infection, tertian and subtertian. Two days after 
administration of the drug temperature became normal. The 5th day blood 
smear became negative. Crescent seen the 9th day, blood thereafter ne- 
gative. After six days of treatment spleen became palpable on inspiration. 


Clinical history, — Fever, headache, and chills for four days regularly 
every afternoon. Spleen two fingers' breadth below left costal margin. 

Treatment. — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until 15th day. 

56,3 Maranon et al.: Totaquina 249 

Observation, — One day after administration of the drug temperature 
became normal. The 5th day blood smear became negative. The 13th day 
spleen could no longer be palpated on inspiration. 


Clinical history, — High fever, headache, and vomiting for two days. 
Spleen not palpable. 

Treatment. — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation, — One day after administration of the drug blood smear be- 
came negative and temperature normal. 


Clinical history, — Fever, headache, and chills every afternoon for three 
days. Spleen not palpable. 

Treatment. — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation. — Two days after administration of the drug temperature 
became normal and blood smear negative the 4th day. 


Clinical history. — Fever, headache, and chills for two days coming every 
afternoon. Spleen tender and palpable on deep inspiration. 

Treatment, — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation,, — Two days after administration of the drug temperature 
became normal and blood smear negative the 5th day. Spleen not palpable 
on day of discharge. 


Clinical history.— ¥ ever, headache, and chills every afternoon for three 
days. Spleen palpable on deep inspiration. 

Treatments — Totaquina, 0.6 gram three times a day for seven days and 
then reduced to 0.3 gram t. i. d. until the 15th day. 

Observation,. — Three days after administration of the drug temperature 
became normal and blood smear negative the 5th day. 


Cases 19 to 21 were all diagnosed as subtertian malaria. Fe- 
ver charts and results of blood-smear examinations are given in 
text fig. 1 and Table 1. Physical examination normal unless 
otherwise stated. 


Clinical history. — Fever, headache, and chills, followed by profuse per- 
spiration every afternoon for three days. Spleen not palpable. 

Treatment. — -Totaquina, 0.6 gram three times a day for nineteen days, 
then reduced to 0.3 gram t. i. d. until discharged two days later. 

Observation, — Two days after administration of the drug temperature 
became normal. Young asexual forms disappeared the 8th day, but game- 
tocytes persisted until the 16th day of treatment, when plasmochin (1 tablet 

250 The Philippine Journal of Science 1935 

t. i. d.) was added to totaquina. Gametocytes then disappeared in forty- 
eight hours. 


Clinical history. — ^Fever, headache, and chilly sensation for six days 
every afternoon. Spleen not palpable. 

Treatment — ^Totaquina, 0.6 gram three times a day. After thirteen days' 
treatment there were still crescent forms in the peripheral blood. Plas- 
mochin simplex then added to treatment, one tablet three times a day. 

Observation, — ^One day after administration of the drug temperature be- 
came normal and young asexual forms disappeared the 6th day. Crescent 
forms appeared in increasing numbers. Two days after adding plasmochin 
to totaquina the blood smear became negative. 


Clinical history,. — Fever occurring irregularly for about two weeks. 
Spleen extends to umbilicus. 

Treatment — Totaquina, 0.2 gram combined with lactose in papers ad- 
ministered with milk three times a day for fourteen days. The 7th day 
plasmochin was added, one tablet once a day for six days. 

Observation. — Three days after administration of the drug, temperature 
became normal and asexual forms disappeared the 6th day. Crescents 
began to appear in increasing numbers. When plasmochin was added to 
treatment, gametocytes began to decrease in number and after four tab- 
lets were given, gametocytes were no longer seen. 


Totaquina in doses of 0.6 gram three times a day usually re- 
lieves the patient of symptoms within two days, at most three 
days. The temperature returns to normal and no more chills 
occur. The schizonts and gametocytes of benign tertian disap- 
pear within the same period, schizonts being cleared first. 

The effect of totaquina on the crescent form of Plasmodium 
falciparum is negligible, just as in quinine or atabrine medica- 
tion, but it is definite on schizonts of the malignant type. As 
was observed, addition of plasmochin to the treatment clears the 
blood of crescents. There were no untoward effects of totaquina 
observed, such as ringing in the ears, vertigo, vomiting, or deaf- 
ness, frequently observed in intense quinine medication. 

We observed that with administration of the drug there was 
a tendency toward increase in the number of gametocytes in 
subtertian cases. This has been observed also with quinine and 
atabrine. It appears that the response of parasites to the drug 
leads to increased formation of gametocytes. 

Whether totaquina would give a thorough cure so as to prevent 
relapses, we are not yet in a position to say, as the period of ob- 

56,3 Maranon et ah: Totaquina 251 

servation after patients were discharged from the hospital was 
too short and the chances of reinfection were too many. 

The fact that totaquina is much less bitter than quinine 
(although more bitter than euquinine) makes it easily admin- 
istered to children, 


This paper briefly reviews the history of cinchona and its 
alkaloids, especially quinine. It points out that cinchona cultiva- 
tion has been found practicable in the Philippines by the Bureau 
of Forestry. It calls attention to the recommendation of a com- 
mittee of the Health Organization, League of Nations, that a 
standard extract of total alkaloids of cinchona be developed as 
a cheap substitute for the purified quinine. The name ''tota- 
quina" was recommended for this new product. 

This paper reports the easy and inexpensive preparation of 
totaquina at the Bureau of Science from Philippine cinchona 
bark. It further reports successful clinical trials with tota- 
quina at the Iwahig Penal Colony Hospital. 

Finally, this paper suggests that the preparation and distribu- 
tion of Philippine totaquina could be developed into a local 
industry of importance, and that it might have a notable effect 
in reducing the incidence of malaria. 


1. It is not improbable that if every case of malaria occurring 
in one year in the Philippines could be given a 250-grain treat- 
ment there would be needed at least 30,000 kilograms more 
specific febrifuge than is now imported. 

2. A 250-grain treatment with quinine sulphate in the prov- 
inces costs from 2.50 to 5 pesos (1.25 to 2.50 dollars United States 
currency) . The greater the need, the higher the price. Quinine 
dihydrochloride retails for from two to four times as much as 
the sulphate. These retail prices are far more than the average 
farmer in the provinces can pay. Quinine and the synthetic 
drugs plasmochin and atabrine may therefore be called a rich 
man's remedies. There is no probability that much more quinine 
can be paid for than is now imported, 

3. The Bureau of Forestry has demonstrated that cinchona 
will grow in the Philippines and will give as good a yield of 
alkaloids as that grown elsewhere. 

4. From our studies we conclude that the standardized total- 
alkaloid-extract of cinchona, recommended by the Health Organ- 

252 The Philippine Journal of Science 1935 

ization of the League of Nations and called "totaquina," can 
be prepared locally from Philippine cinchona easily and inex- 

5. We conclude from some clinical tests that this Philippine 
totaquina is probably about equal to quinine sulphate in its 
therapeutic value against malaria. We are not able to form any 
conclusion as regards relapses, but we found that totaquina 
subdues the acute attack effectively and without untoward effects. 
It does not destroy the crescents of subtertian malaria, being 
similar to quinine and atabrine in this respect. 

6. We conclude that, allowing a fair profit to the grower of 
cinchona, the manufacturer of totaquina, and to the retailer, 
this Philippine totaquina could be sold to the people at not more 
than 35 centavos (0.175 dollar) per 250-grain treatment. 

7. Contrasting 35 centavos with the present retail price of 
from 2.50 to 5 pesos for a 250-grain treatment with quinine 
sulphate or atabrine, we conclude that the local production of 
totaquina would materially aid in combating malaria in the 

8. We also conclude that the growing of cinchona and the 
manufacture of totaquina might have considerable economic im- 
portance to the Islands, being capable of becoming sizeable new 



Cost of 5,016 seedlings to be planted in 1 hectare with 
spacing of 1.5 by 1.5 meters at 11 pesos per 1,000 55.18 

Average cost of clearing, preparing, and planting 1 
hectare 114.00 

Total 169.18 

Interest due 169.18 pesos for seven years at 7 per cent, 
compounded annually 102.52 

Annual expenses for up-keep and supervision, ex- 
cluding land rent and taxes at 127 pesos per hec- 
tare, plus the interest on the amount at 7 per cent, 
compounded annually 1,099.46 

Cost of harvesting and curing the bark of 2,721 trees 
at 0.10 centavos per tree 272.10 

Total expenses „ 1,643.26 

Yield of 2,721 trees at 0.96 kg per tree, 2,612 kg at 

0.87 centavos per kg. 2,272.53 

Profit at the end of the seventh year. 629.32 

Income per hectare per year 89.90 

56,3 Maranon et al: Totaquina 253 


1. League of Nations, Health Organization. Conclusions of the Sub- 

committee of Experts on Quinetum. C. H. (Malaria) 167 (1931) 2 
pp. mimeographed. See also mimeographed report C. H. (Malaria) 
183 Geneva (October 10, 1932). 

2. Russell, P. F. Malaria and Culicidae in the Philippine Islands; his- 

tory and critical bibliography, 1898 to 1933. Technical bulletin 1. 
Dept. Agriculture and Commerce, Manila (1934) 62, 63. Tables 3 
and 4. 

3. League of Nations, Health Organization. Enquiry into the quinine 

requirements of malarial countries and the world prevalence of ma- 
laria. C. H. (Malaria) 185 Geneva (1932) 1-89. 

4. SUPPAN, Leo. Three Centuries of Cinchona. Proc. Celebration 300th 

Anniversary First Recognized Use of Cinchona. St. Louis (1931) 

5. Fischer, A. F., and A. L. Cenabrb. Possibility and financial aspects 

of growing cinchoma in the Philippines. (Unpublished.) 
Fischer, A. F. Forest Planting. Cinchona in Bukidnon. Memoran- 
dum (1931). (Unpublished.) 

6. Wester, P. J. Notes on cinchona in Java. Philip. Agr. Rev. 9 (1916) 


7. Anonymous. Cinchona experiments in the Philippines. Footnote. 

Chemist and Druggist 91 (1919) 45. 

8. Altamirano, G. M. Cinchona (quinine) cultivation in the Philippines. 

Manuscript in library of Bu. Forestry, Manila (1934) 7 pp. type- 
written. (Unpublished.) 

9. Bu. Forestry, Manila. World market exists for Philippine quinine. 

The Tribune (March 31, 1934). 

10. Malaria Commission. 3d General Report. Quarterly Bulletin. Health 

Organ. League of Nations, Geneva 2 (1933) 181-285. 

11. Howard, B. F. Some notes on the cinchona industry. Chemical News 

142 (1931) 129-133. 

12. GiEMSA, G. Etude comparative d'^chantillons de totaquina, de quinine 

et d^hydroquinine en point de vue de la toxicite et de TefRcacite dans 
la paludisme des oiseaux. Riv. di Mai. 12 (1933) 70-86. See also 
G. Giemsa and M. Oesterlin. Beihefte z. Arch. f. Schiffs- u. Trop.- 
Hyg. 37 (1933) 217-244. 

13. Groothoff, A., and T. A. Henry. The preparation, analysis, and 

standardization of totaquina. Riv. di Mai. 12 (1933) 87-91. 

14. The British Pharmacopoeia 1932. Constable & Co. Ltd., London (1932) 


15. Dawson, W. T. Antimalarial value of cinchona alkaloids other than 

quinine. South. Med. Journ. 25 (1932) 529-533. See also Journ. 
Roy. Army Med. Corps (March, 1931). 

16. Henry, T. A. The plant alkaloids. 2d ed. P. Blakiston's Son & Co. 

Phila. (1924). 

17. Fletcher, Wm. Treatment of malaria with the alkaloids of cinchona. 

Inst. Med. Res. Kuala Lumpur (1923) 1-91; (1925) 1-19. 

18. Field, J. W. Notes on "Totaquina." Malayan Med. Journ. 9 (1934) 


254 The Philippine Journal of Science 

19. SiNTON, J. A. The standardization of mixed preparations of the cin- 

chona alkaloids in relation to Indian conditions. Rec. Malaria Surv. 
India 4 (1934) 5-13. 

20. Howard, D., and 0. Chick. Thorpe's Dictionary of Applied Chemis- 

try. 2d ed. 2 (1927) 260-261. 

21. Pharmaceutical Journal and Transactions III 12 (1881-82) 662. 

22. Anonymous. Jardin Botanico. La Ilustracion Filipina, Manila 3 

(1893) 75. 



Fig. 1. Temperature charts, totaquina therapy, Iwahig Penal Colony Hos- 
pital. Patients were hospitalized for at least fifteen days. The 
part of the temperature charts not shown above presented only 
normal temperatures. 
2. Totaquina therapy, Iwahig Penal Colony Hospital. Blood-smear 



By Paul F. Russell 
Of the International Health Division of the Rockefeller Foundation 


Francisco E. Baisas 

Of the Philippine Bureau of Health, Manila 


The authors have recently prepared a series of papers to 
facilitate the practical identification of Philippine Anopheles Ah 
2,3) This discussion of technic is added to supplement those 
papers, in the hope that it will increase their usefulness. It 
is intended to be simple enough for ordinary field use in the 
Islands and yet sufficiently comprehensive to assist those who 
may desire to do more than merely identify a given larva or 

We have found the following references of particular value 
with regard to the technic of handling mosquitoes: Boyd, (4) 
Christophers, (5) and Gater.(6) Many other references could 
be given. Most general books on tropical medicine include a 
discussion of this subject. 

We cannot resist stating at the very outset that, to be suc- 
cessful, an inspector or other officer interested in malaria re- 
quires not so much finesse in technic as curiosity, enthusiasm, 
and determination. Field work in malariology will not permit 
white collars or polished shoes. It is frequently arduous, dirty, 
and damp. 

Furthermore, there are almost as many technical methods of 
dealing with mosquitoes as there are malariologists. We can 
only present those which have seemed best to us. They should 

* This paper has been prepared as a part of the program of Malaria In- 
vestigations, a project which has been jointly supported by the Bureau of 
Science, Manila, and the International Health Division of the Rocke- 
feller Foundation. The Bureau of Health has cooperated through the as- 
signment of the junior author as chief field director. 


258 The Philippine Journal of Science i9S5 

be modified and improved to suit individual needs and not 
followed slavishly. 


Individual anopheline or aedine mosquito eggs are so small 
that they are rarely found in natural breeding places. Cidex 
mosquitoes lay their ova in rafts (text fig. 1), and these are 
easily seen floating on the surface of stagnant water; but Ano- 
pheles ova are usually deposited singly and escape detection (text 
fig. 2). 

Ova for study may be obtained from gravid females. When 
first laid they are white, but they soon become black. They vary 
in length from 0.6 to 0.8 millimeter, and each species has definite 
morphological characters. For some countries keys to the Ano- 
pheles ova have been prepared. For example, in Europe six 
varieties of A, maculipennis have been identified by their 

Fro. 1. Cidex egg raft. 

eggs. (7) It would be well to study the ova of Philippine Ano- 
pheles, a subject on which nothing has been published. Differ- 
entiation is made on the basis of size, shape, and position of the 
lateral floats, as well as the design on the dorsal surface of 
the ovum. Color may also be important. (See text fig. 2.) 

According to Hackett(7) anophelines with ripe eggs will 
readily lay them in small cotton-stoppered vials, 50 by 20 milli- 
meters, on wet filter paper or pressed cotton. 'The eggs are 
usually laid in a small heap and they can be dispersed and at 
the same time preserved by dropping 2 per cent formalin on 
them from a pipette or dropping bottle (five or six drops). 
The eggs can be kept on wet cotton in small entomologist's tin 
specimen boxes, IJ in. in diameter, which have a glass set in 
the lid, through which the eggs can be examined. Washers or 
rubber rings cut from tubing the same diameter as the tin box 
serve to keep the cotton at the bottom from falling against 
the lid if the box should be overturned or sent by post." We 
have not tried this method. 


Riissell and Baiscis: Handling Mosquitoes 


Fig. 2. Two types of egga of Anaphelea minimus var. flaviroatris. 



The following equipment is essential for collecting mosquito 
larvse : 

1. Dipper. — Any sort of a receptacle may be used, from a co- 
conut shell to a standardized copper dipper specially made for 
mosquito surveys. It is best as a rule to use a tin or enamel 
dipper having a long hollow handle into which a cane may be 
thrust. In some places a shallow pie tin or deeper cake tin may 
be useful. In other places, such as shallow water, a large spoon 
is most serviceable. If comparative quantitative studies are 
being made, a standardized dipper of about 400 cubic centimeters 
capacity should be used. 

260 The Philippine Journal of Science 1935 

We have found very useful an Italian type of tin dipper having 
a screen window protected by an outer shell (text fig. 3) . This 

screen permits water to be 
poured out of the dipper with- 
out loss of larvse and thus fa- 
cilitates the collection. If the 
inside of the dipper is painted 
white it is easier to see the lar- 

2. Pipette or spoon for remov- 
ing larvse from the dipper. — An 
Fig. 8. Special type of dipper. Ordinary medicine dropper with 

the narrow end cut off is satis- 
factory, although wider special collecting pipettes may be pur- 
chased and are more desirable. Narrow medicine droppers 
are not suitable for removing large culicine larvse or pupse, in 
fact these may be killed when drawn into the ordinary pipette. 
In such cases, if no large pipette is available it is better to use 
a spoon. 

S. Collection bottle. — Some sort of a collection bottle is needed 
in which to carry the larvse from the field to the laboratory. 
Any wide-mouthed bottle that can be stoppered will do. We 
have found half -pint mason jars satisfactory. These can be 
fastened to the belt. 

4. Carrying case. — A carrying case is required for the bottles. 
We have used a simple wooden case with a rack to support six 
bottles upright (Plate 1, fig. 3). An army bag is useful in 
the mountains or difficult terrain. It can be slung over the shoul- 
ders or strapped to the back and permits the hands to remain 

5. Notebook. — It is essential to keep an adequate record of 
the collections. A notebook and a pencil should be carried in 
the field so that notes can be made at the breeding place. 

6. Bolo. — It is always wise to carry a bolo or other large 
knife to cut undergrowth and to kill snakes. 

There are some practical points about larva collecting which 
should be noted. In the first place the type of breeding place 
to be searched is of importance. If the object of collecting is 
to make a complete anopheline and culicine survey, no accumula- 
tion of water can safely be overlooked. Even tree holes and 
miscellaneous rain-catching hollows and receptacles must be 
examined. Aedes mosquitoes will breed in a gill of water, even 
within houses, as in ant guards and flower pots. 

56,3 Russell and Baisas: Handling Mosquitoes 261 

If only anophelines are wanted, the following types of breed- 
ing places may be included in the survey: 

Running water in creeks, streams, rivers, ditches, canals, and 

Stagnant fresh water in wells, lakes, ponds, pools, ditches, and 
swamps, as well as in tree holes, rock pools, temporary rain 
puddles and hoof prints, and in miscellaneous barrels and tubs. 
Rice fields are a prolific source of Anopheles mosquitoes. 

Salt water in salt ponds, fishponds, lagoons, swamps, and 
river mouths. 

Finally, if the object is a malaria survey, collections may be 
limited to running water, including springs and spring-fed 

The proven malaria-carriers in the Philippines are A. minimus 
var. flavirostris and A. maculatus. The former is of paramount 
importance, while the latter has rarely been incriminated. 
Both species breed in running water, and neither has been taken 
in either salt water or completely stagnant water, except in fresh 
residual overflow pools in stream beds. 

Anopheles minimus var. flavirostris is most abundant in 
foothill streams, close to partially shaded edges, especially where 
there are bamboo roots. It has not been found above 2,000 
feet altitude. Anopheles maculatus has been taken up to 5,000 
feet altitude. It is most abundant along the edges of partially 
shaded forest streams. 

In collecting it is best to approach the actual breeding place 
without disturbing it, for the larvse are easily alarmed and 
quickly submerge or scatter. They can remain submerged for 
several minutes. Therefore, while wading in the stream, the 
collector should try not to send waves toward the breeding 
place. If feasible, he should stand on the bank while dipping. 
The dipper should not hover over a breeding place but make de- 
cisive, clean-cut dips. It is to be remembered that the anopheline 
larvae are on the surface, and surface water should therefore get 
into the dipper as expeditiously as possible. If dipping is done 
where there are roots, the dipper should be so manipulated that 
surface water will rush into it from among the roots, carrying 
with it the larvae. It may be necessary to dip several times in 
one spot before the larvae are taken. 

In contrast to this "sucking" maneuver in which the water 
is drawn out from among roots, one may in some places use a 
sweeping movement of the dipper, skimming a considerable area 

290499 3 

262 The Philippine Journal of Science 1935 

of the surface of the breeding place in one arclike sweep. 
When there is debris or a floating log, the dipper may be quietly 
submerged beside it so that currents of water will carry the 
larvse into the dipper, somewhat as in the case of larvse from 
roots, described above. If the dipper overflows, many larvae 
may be lost. If the water is turbid or full of debris, it may take 
several minutes to remove the larvse from the dipper, as they 
are prone to remain for a time at the bottom. It is more difficult 
to take pupae than larvae, the former being quicker to see danger 
and to escape from it. 

Small puddles and hoof prints may be stirred first, before 
dipping. The muddy water will furnish a background against 
which the larvae, after a little time, are easily seen and removed 
directly with a pipette or spoon. For tree holes a long pipette 
is useful. One may easily be made from glass tubing and a 
large rubber bulb. If the tree hole is dry some of the mate- 
rial in the bottom should be taken. The ova of tree-hole mos- 
quitoes resist drying, and sometimes, when the removed material 
is put in water in the laboratory, larvae will appear. 

For wells the dipper used by Russell and Santiago (8) is very 
useful. It has a long handle with a hinge and set screw near 
the dipper so that it may be given various angles to adapt it 
to existing circumstances. 

The capture of only a few first-stage larvae does not seriously 
incriminate a breeding place for it does not prove that larva 
will grow to maturity in the given site. But if large larvae or 
pupae are taken, or if larval skins are seen, then it may be 
assumed that the breeding place is of importance for the species 

If ten adequate dips along a meter of stream bank are fruit- 
less, that site may safely be called negative at the time of exam- 

Negative results may be due to inexperience in collecting, to 
recent disturbances, such as flushing, abrupt change in water 
level, agitation, of the breeding place, application of larvicide, or 
to other causes. Therefore, a place found negative should not 
be omitted on subsequent visits, but subjected to repeated 

Field notes should include not only the date and the general 
type of breeding place, such as stream, pool, or spring, but also 
a statement about shade — ^whether dense, medium, light, or 
absent. It is well also to record whether or not the water is 
clear or muddy, what type of bottom, what type of vegetation 

56r3 Russell and Baisas: Handling Mosquitoes 263 

is along the bank, what water plants are present, and whether 
natural enemies, such as top minnows or dragon-fly larvse, are 

If quantitative studies are being made, the larva of different 
stages should be counted separately and the number of dips 
recorded. In all cases the place of collection should be carefully 
stated; that is, name of area, kilometer post, or distance and 
direction from a fixed, well-known point. If a detailed survey 
is being made it is essential to prepare a sketch map of the area, 
including roads, settlements, and bodies of water. 

Finally, collecting larvse successfully requires considerable 
practice. Many a place has been called negative because the 
would-be collector was not adept. The dry-foot man is never 
to be trusted because it is utterly impossible to make adequate 
collections without frequently entering the water. Experienced 
collectors sometimes rely on the appearance of a place and dip 
only in the likely places. This is dangerous, as one can be easily 
fooled. Larvae are sometimes found in abnormal sites, especially 
in a dry season. 


Anopheline larvae can be transported considerable distances, 
if reasonable care is exercised. We have found that the half- 
pint mason jars mentioned above are suitable. Several may be 
moved about in a wooden (not metal) collecting box which 
has a rack inside to keep the bottles upright, and a handle to 
facilitate carrying. If the distance to be covered is not great 
and the road is fairly smooth, the bottles may be two-thirds full 
of water with perhaps a small amount of green algae added to 
minimize the motion of the water. If the larva must be trans- 
ported over longer distances, it is well not to have the bottles 
more than a quarter full, and it is essential to add algse. The 
bottles must be correctly labeled, and no natural enemies should 
be included in them. If jars are not available, bamboo joints 
may be used satisfactorily. They are especially useful for larvae 
from rock holes and artificial containers. We have transported 
living larvae in such bamboo joints from Mindanao to Manila, 
a four days' journey. A good number of leucosphyrus and near- 
leucosphyrus emerged from this collection. The top of the con- 
tainer was covered with gauze held in place by rubber bands. 

Sometimes living larvae may be sent through the mails in 
a special type of bottle, stoppered, but having a glass tube pierc- 
ing the cork and extending about halfway into the bottle. The 


The Philippine Journal of Science 


quantity of water used is such that, no matter what the posi- 
tion of the bottle is, the surface of the water does not touch the 
open end of the tube (text fig. 4). 

Boyd (4) states that for transportation over very rough ground 
or on horseback it is a good plan to transfer the larvse to a bottle 

Fig. 4. Larva shipping bottle. 

which has been freshly filled with water and emptied, so that 
the inside is wet. The larvae become deposited nearly immobil- 
ized in the film of moisture and the surplus water is discarded. 
The bottle is then tightly sealed. A second bottle is filled with 
water from the breeding place, and this furnishes a supply to 
be used in the laboratory for rearing the larvse. 

56,3 RiLSsell and Balsas: Handling Mosquitoes 265 


When the larvae have been taken to the laboratory they 
should be transferred to properly numbered or labeled bowls 
or pans, which have more surface area than the bottles. Spe- 
cial attention must be paid to light, temperature, and food. For 
proper development the larvse require some direct sunlight, but 
in the Tropics it is sufficient to expose them only from about 6 
to 7 a. m. For the rest of the day it is better not to have them 
directly in the sun. An appreciable rise in the temperature of 
the water above normal will result in high mortality. 

Numerous foods have been recommended for larvae. We have 
found a mixture of Loffler's dehydrated blood serum and litmus 
milk satisfactory for all species of mosquito larvse. Very small 
amounts are used. Another good food is yeast. A piece about 
half a centimeter square may be put in the bowl on a support 
of some kind so that the yeast is held about a centimeter or so 
below the surface of the water. 

Where these special foods are not available good results may 
be had with natural algae, finely chopped cockroaches, or hay 
infusion. The last is made simply by adding some hay or rice 
straw to water and allowing it to soak. Algae from salt-water 
ponds, boiled and allowed to stand for a few days, form a white 
scum which is an excellent food for larvae. It is somewhat 
odorous and should be protected by gauze from flies. Regardless 
of what food is used the water must not be allowed to get foul 
and should be changed as necessary. 

It has seemed to us that aerating the water accelerates devel- 
opment of the larvae. This is done simply by means of a pipette, 
repeatedly but gently injecting air below the surface of the 
water in the bowl. 

It is sometimes a good plan to separate the various stages of 
larvae, for the older may attack and devour the younger ones. 
Culicine larvae should be kept separately as many species are 
cannibalistic. As a general rule, unless the species is a rare 
one, it is best to discard the small larvae and to rear only those 
of the third and fourth stage. The smaller ones require much 
attention and give unsatisfactory results. 

For rearing large numbers of A. minimus var. flavirostris lar- 
vae Russell and Santiago (9) erected a wire cage in a stream at the 
bank. The water of the stream flowed through a very fine 
screen at each end of the cage, simulating natural conditions. 

266 The Philippine Journal of Science 1935 

In the laboratory a flow of water through a basin may be 
secured from a receptacle used as a reservoir above the basin 
and connected with it by a thick thread of lampwick. This wick 
or thread extends from the water in the reservoir down to and 
through the basin containing the larvse, and hangs over the edge 
of this basin. Water flows by capillary attraction and gravity 
down the thread to the basin and from there to waste. 

It is necessary to inspect the bowls of larvse each day. Pupae 
are transferred to individual vials or test tubes about one-third 
full of water, and lightly plugged at the top with cotton. Pupae 
do not feed, but are very susceptible to high temperatures and 
to rough handling. A small twig may be put in the vial or test 
tube as a resting place for the adult when it emerges. The vials 
or test tubes must be properly numbered or labeled. 

Adults should be transferred to cages, lantern chimneys, or 
dry, clean vials, also properly labeled. Leaving them in the 
vial with water usually results in destruction of fringe spots and 
scales of the wings. But this transfer should not be made for 
about six hours, so that the adult will have time to dry out. 

It is frequently a good plan to save the last skin, or pelt, 
which is molted when the larva becomes a pupa. This pelt is 
preserved, as noted below, and affords a very useful check on 
the identification when its identification is compared with that 
of the adult which emerges from the pupa. Making a diagnosis 
from both the larval skin and the adult avoids any possibility 
of confusion. 

Each collector will by experience find out the best way to rear 
his larvse under the conditions prevailing in his own laboratory. 
Surprisingly good results can be had with makeshift equipment, 
provided the collector has sufficient interest in the subject, 


It is always necessary to keep in mind the fact that the iden- 
tification of larvse or pupae depends for the most part on minute 
body hairs. These are fragile and are easily rubbed off by 
rough handling. Therefore, the specimens must be handled 
with delicacy. It is a good rule to manipulate them singly and 
not to attempt to save time by carrying several in one manipula- 
tion of a pipette. 

For temporary preservation the larvae or pupae may be killed 
in hot, but not boiling water (temperature about 65° C.) . They 
may be transferred from the collecting bottle to the hot water 
and removed as soon as dead. Leaving them longer than neces- 

8».8 Russell and Baisas: Handling Mosquitoes 267 

sary in the hot water tends to bend the bodies and to make 
further examination or manipulation difficult. They may then 
be picked up by a pipette and put into a 10 per cent solution 
of acetic acid where they will retain their characters for a few 
days, at least. If chloral hydrate is available a more suitable 
temporary preservative is made by dissolving 57 grams of 
chloral hydrate in water and adding 6 cubic centimeters of gla- 
cial acetic acid and enough water to make 100 cubic centimeters 
of the preserving fluid. 

Either of these temporary preserving fluids is suitable for 
forwarding specimens from the provinces to Manila, but neither 
is advisable for permanent preservation. 

If permanent preservation of unmounted larvae or pupae is 
desired, MacGregor's solution is suitable. This is made by 
dissolving 5 grams of borax in water, and adding 2.5 cubic cen- 
timeters of glycerin, 100 cubic centimeters of a 40 per cent so- 
lution of formaldehyde, and enough distilled water to make 1 
liter. After thorough mixing the solution is filtered. Larvae or 
pupae, after being killed in hot water, are placed in this solu- 
tion, in small vials or bottles tightly corked and sealed with 
wax or paraffin. It is always wise to put a pencil-written label 
inside the vial and to add another label, written in India ink, 
on the outside of the vial. If this outside label is brushed with 
Bureau of Science book varnish it will be impervious to weather- 
ing and to cockroaches. We have found MacGregor's solution 
very satisfactory. It keeps the larvae flexible, whereas in 10 
per cent formalin they tend to become bent and somewhat brittle. 

If considerable numbers of larvae are to be preserved in one 
vial it is better to make a double transfer from water to a 
small beaker of the preservative and thence to a vial of fresh 
preservative. This prevents undue dilution. Larvae and pupae 
so mounted will keep their characters for many years, even 
under tropical conditions. 

Larval pelts may be preserved temporarily in 70 per cent 
alcohol in individual vials, properly numbered. We prefer 
the borax-formalin solution described above. For permanent 
preservation it is best to mount the pelt, as described below. 


Larvae and pupae may be mounted for temporary or perma- 
nent use. Temporary mounts may be made of either living or 
preserved specimens. Living larvae or pupae may be mounted in 
a drop of water on a glass slide for diagnosis. If they are 

268 The Philippine Journal of Science 1935 

handled with reasonable care they may be returned to the bowls 
and reared into adults. If it is necessary to use a cover slip 
on a living larva it is better to put a small piece of paper or 
broken cover glass, or a hair, under the slip to avoid injury 
to the specimen. The careful use of ice water or of weak chloro- 
form water will tend to immobilize the larva without killing it. 

Preserved larvae may be mounted temporarily for examination 
in a drop of the preserving solution. It is usually better to 
use a cover slip and to mount the larvae dorsum uppermost near 
the middle of the slide. 

Permanent mounts give more satisfaction than the temporary 
ones. They may be made from either living or preserved larvae 
or pupae. We prefer Gater's(6) method of mounting. Based 
on his description the process for living larvae is as follows : 

1. Prepare the mounting fluid. This is done by placing 8 grams of 
picked gum arabic in 10 cubic centimeters of distilled water in a beaker 
which is covered and kept in a warm, but not hot, place, such as an in- 
cubator, stirring it occasionally until it is dissolved. After it has dis- 
solved add 74 grams of chloral hydrate, stir and replace in incubator 
until dissolved. Then add 5 cubic centimeters of glucose syrup (98 grams 
of bacteriological glucose dissolved in 100 cubic centimeters distilled wa- 
ter) and 3 cubic centimeters glacial acetic acid. Stir and allow to stand. 
Then filter through No. 5 Whatman paper in a Buchner funnel, using a 
suction pump. If this is impossible, allow the solution to stand until all 
sediment has settled and then decant the upper portion. [If this solution 
cannot be made in a local laboratory a small supply may be obtained from 
the entomologist at the Bureau of Science or from the malaria control sec- 
tion of the Bureau of Health, Manila.] 

2. Transfer the larva from the collecting bottle or rearing bowl to 
the center of a glass slide. We have found it a good plan to trace the 
outline of a standard slide on a piece of cardboard and on this trace 
the outline of a cover glass, marking the center. This furnishes a guide 
for mounting. The specimen should be placed with the head toward the 
one doing the mounting. A little practice will soon result in perfectly 
aligned slides so that there will be complete uniformity of appearance 
in a box of slide mounts. 

3. Remove as much water as possible, first with a capillary pipette and 
then with a piece of blotting paper until the specimen is almost dry. Avoid 
touching the larva as much as possible. If it gets out of position it can 
be maneuvered into place by tapping and tilting the slide. Experienced tech- 
nicians, however, can arrange and rearrange the larva and its hairs using 
very fine dissecting pins and a binocular to get the best results. Allow 
the larva to adhere to the glass so that it will remain in place during 
subsequent manipulations. 

4. With a fine glass rod place a drop of the mounting medium, de- 
scribed in paragraph 1 above, over the larva as it lies with dorsal part 
of head and body uppermost. When the larva is dead add another drop 

56, a Russell and Baisas: Handling Mosquitoes 269 

of mounting fluid and carefully place a clean cover glass over the spe- 
cimen. The fluid decreases in volume as it dries so an excess should be 
used. This medium sets hard in about a week but the larva will be 
cleared and suitable for examination in a few hours. 

Slightly better results may be had if the larva is killed in hot water 
and then kept overnight in the chloral hydrate medium described under 
temporary preservation above. 

5. Add the proper label at the right end of the slide as the head of 
the specimen is towards the one doing the mounting. Brush the label 
with a collodion or book varnish to preserve it. 

6. When the mounting medium is dry the cover glass should be ringed 
as neatly as possible with a cellulose varnish such as Duco. This is best 
done with a turntable but may be accomplished satisfactorily by hand. 

Mounting larvae which have been preserved is done in the same 
way as just described, except that, if MacGregor's solution is 
the preservative, it is advisable on removing the larva to place 
it first in water for an hour and then overnight in the chloral 
hydrate temporary solution described above. Larvae which have 
been temporarily preserved in acetic acid may be directly 
mounted but are better if given a preliminary bath in this 
chloral hydrate solution. 

The arrangement of the specimen on the slide so that the 
hairs are properly disposed requires a careful technic, which 
comes only after considerable practice. It is essential to use 
just the right amount of mounting fluid as larger or smaller 
drops will spoil the mounting. We have found satisfactory three 
drops from an ordinary toothpick when the smaller end of the 
pick is dipped for about one-third of its length into the fluid, 
and the initial drop is allowed to fall into the container as the 
toothpick is removed. Care should be taken that the fluid 
spreads evenly over the specimen which should be entirely sub- 
merged. The cover slip is dropped squarely on the top of the 

The use of forceps in placing the cover glass may not be 
satisfactory as the glass may easily slip out of position. Thumb 
and forefinger are usually better tools. 

Larval pelts may be mounted temporarily in a drop of the 
70 per cent alcohol preservative. Permanent mounts of these 
pelts are extremely difficult and require considerable practice. 
Gater(6) gives detailed instructions which we are not including 
here as we suggest that, if permanent mounts are desired, the 
pelts be forwarded in the preserving fluid either to the entomolo- 
gist of the Bureau of Science or to the malaria control section of 
the Bureau of Health, Manila. 

270 The Philippine Journal of Science 1935 


For the identification of the larvae of Philippine Anopheles 
we have recently prepared a wall chart, distributed by the 
Bureau of Health, and an illustrated key, (2) which should be 
consulted. Other papers of importance in this regard are those 
by Manalang,(lO) Baisas,(ii) and King. (12) The pupal charac- 
ters of Philippine species are being studied by the junior author. 

Before attempting to identify a larva it is, of course, necessary 
to be acquainted with the larval characters used in a key. We 
have described this subject at length in our key. (2) 

Having become familiar with the larval characters, one should 
proceed with the identification systematically. Head, thorax, 
and abdomen should be examined in order and the characteris- 
tics of the important hairs, the tergal plates, and the pecten 
noted. We prefer to use the low power of a compound micro- 
scope in the identification of larvse. With a little practice it be- 
comes possible to recognize all of the essential characters. One 
of us (13) has used a microscope mounted on a tripod by Darling. 
This was taken into the field and the species of larvse in a given 
breeding place were identified at once. 

When the characters have been studied it becomes possible, 
by means of the key, to make an identification ; but usually more 
than one character is required for positive diagnosis. There are 
many individual variations. Furthermore, although Barber et 
al.,(i4) Holt and Russell, (15) and Russell (16) have published 
fairly comprehensive reconnaissance studies, there has not been 
an adequate survey of Philippine Anopheles. Therefore, the 
possibility of encountering a new species or variety is always 
present. If a careful study of a specimen fails to establish a 
clear diagnosis, it should be sent to the malaria control section 
of the Bureau of Health. 


It is sometimes desirable to dissect larvse to determine their 
diet or to observe whether or not they have ingested a larvicide 
mixture, such as Paris green and charcoal. 

Apparently, most anopheline larvae ingest a heterogeneous diet, 
consisting of any sort of inert or living material small enough 
for passage through their gullets; but their nutrition seems to 
depend on the digestion of such green algae as Euglena and 
Cosmarium, diatoms, rotifers, ciliates, and other protozoa. No 
studies have been reported on the diet of Philippine Anopheles 
larvse, although it is a matter of importance. 

5«,8 Russell and Baisas: Handling Mosquitoes 271 

The dissection of a larva offers little difficulty. With dissect- 
ing needles the anterior part of the thorax is cut off. Then the 
abdomen is nicked just anterior to the last segment and the 
gut attached to this terminal segment drawn out. The gut is 
transparent, and if it is flattened in a drop of water or saline 
solution under a cover slip, such contents as Paris green, char- 
coal, or protozoa can readily be distinguished. 

As noted by Boyd and Foot(l7) the digestion of a larva pro- 
ceeds very rapidly, and owing to the disturbance of the water 
in a collecting bottle the larva may not feed during transit. 
For this reason the gut may be nearly empty by the time the 
larva is examined in a laboratory. Therefore, the specimen 
for examination of gut contents should at the time of capture 
be put into a solution of 10 per cent formalin 9 parts, and 
glycerin 1 part. It is better to use only fourth-stage larvse. 
Then, in the laboratory, after making an identification of species, 
the gut may be removed and crushed in a little water on a glass 
slide. Several specimens from larvse of the same species may 
be put together and a microscopical analysis of the total gut 
contents made. 

In larval food studies it is usual also to collect about a half 
pint of water from the breeding place, and, after shaking and 
centrifuging in the laboratory, to examine the sediment for 



The following equipment is essential for collecting adult mos- 
quitoes (Plate 3, figs. 1 to 3). 

Flashlight— Adult mosquitoes usually are caught either at 
night or in dark places in the daytime. Therefore, a flashlight 
is indispensable. An ordinary two- or three-cell dry-battery 
light is suitable. If the light can be focused it may be slightly 
more convenient, but this is not essential. A supply of batteries 
and bulbs should be available. Incidentally, as noted below, a 
powerful flashlight is very useful in the laboratory for the exam- 
ination of adults. 

Catching tube, — Many types of catching tubes are used. The 
simplest and the one which has given us good results is a small 
vial measuring about 8 by 1.5 centimeters. Test tubes may be 
used. With a little skill such a vial or tube can be placed over 
a resting anopheline and quickly corked with the disturbed 
insect inside. Only one insect should be kept in a vial. 

272 The Philippine JouTTial of Science 1935 

Another very useful collecting tube is easily made by attach- 
ing a length of rubber tubing over one end of a glass tube. 
The tube should have a diameter of about 8 to 10 millimeters 
and a length of about 30 centimeters. The rubber tubing is 
about 45 centimeters long. The end of the glass tube over which 
the rubber is tied should be first covered by a piece of gauze. 
Suction is made by the mouth at the end of the rubber tubing 
as the free end of the glass is placed near a mosquito, which 
is easily sucked into the tube and then gently blown into a 
collecting vial or killing bottle. Large numbers of insects can 
be caught by this tube in a very short time. Even flying mos- 
quitoes can sometimes be caught. This method is somewhat 
rougher than hand-catching with individual vials, but, with 
practice, it is satisfactory. 

Various types of special catching tubes are sold by entomo- 
logical supply houses. One which has had widespread use is a 
thick-walled, hollow, glass cylinder about 15 centimeters long 
and 3 centimeters in diameter. It has one inverted conical end 
which is perforated by a small hole. The other end is covered 
with a piece of gauze and corked. Through a hole in the cork 
there is a short glass tube on the outer end of which is a rubber 
suction bulb. Mosquitoes are sucked into this tube and cannot 
easily escape. 

Carrying box or case to transport the insects to the labor- 
atory. — Any suitable box or bag or even the collector's pockets 
may be used to carry the collecting vials or tubes of mosquitoes. 
It must be remembered that exposure to direct sunlight quickly 
kills mosquitoes, as does prolonged drying or jarring. 

Notebook. — It is essential to label the catch properly and to 
keep a field record of the collecting places. If quantitative 
studies are being made, the length of time spent in searching 
should be recorded. 

It is very unusual to catch Anopheles adults in, on, or under 
Filipino nipa houses in the daytime. These insects enter such 
houses freely at night and when blooded have been observed to 
linger inside until daybreak, but when it becomes light the 
Anopheles mosquitoes seek other daytime shelter. Nipa houses 
are light, dry, and airy, whereas the insect prefers to rest where 
it is dark, damp, and quiet. Typical daytime resting places for 
local anophelines are small undercut caves along stream banks, 
well-shaded mossy stone walls such as those near old churches 
or cemeteries, small sheltered cuts, or the interstices of the inner 

96,3 Russell and Baisds: Handling Mosquitoes 273 

wall of a shallow well, on the moist walls of partially covered 
cisterns, under cement structures, and sometimes inside strong- 
material houses in a dark, damp room. Adult anophelines may 
be caught by day or by night and will be discussed under these 
two headings. 


Philippine Anopheles have a flight range of at least 2 kilo- 
meters. They are not active by day but only at night, and 
their daylight hours are spent quietly in such shelters as de- 
scribed above. Therefore, if daytime catches are to be made, 
typical shelters must be searched. This presents no difficulties, 
except that occasionally snakes will be found in the same 
small caves chosen by the mosquitoes. An efficient way to 
collect anophelines by day is to walk down a stream known 
to be a breeding place, methodically searching each undercut 
with a flashlight. Not only the stream-breeding species will be 
found, but also others from nearby rice fields and pools. If 
there are well-shaded old stone walls in the vicinity they may 
also be a fruitful source of anophelines. 

Artificial daytime shelters can be made by lining boxes with 
moist earth. The earth-lined box traps that we have used 
measured about 1 by 0.5 by 0.5 meter. (2i) The bottom and 
one end of the box are removed so that the floor of the trap 
is earth. With wire screening to hold it, a layer of moist earth 
about 1 inch thick is applied to the inside of the remaining three 
sides and the top. The open end is partly closed with a black 
cloth, which hangs to within 5 or 10 centimeters of the ground. 
This provides the insect with a dark, damp, quiet, and earthy 
resting place, such as it naturally seeks. Such traps may be 
placed near breeding places, under houses, or wherever it is 
desired to trap anophelines. The insects are easily taken with 
any of the usual collecting tubes, as they rest quietly on the 


Anophelines in the Philippines may be caught at night on the 
walls of bedrooms, on mosquito nets, or feeding on carabaos or 
cows. Many local species, including the malaria-carrying A. 
minimus var. flavirostris, will feed on carabaos, and can be 
easily caught while doing so. By the use of a fairly tame beast 
and a flashlight good catches can be made with any of the usual 
collecting tubes from about 7.30 p. m. until just before daylight. 

274 The Philippine Journal of Science i936 

We have had success with animal-baited traps. These are 
screened structures of various sizes, usually made somewhat 
larger than is necessary to accommodate one carabao. The cara- 
bao is tethered inside, surrounded by a bamboo railing to prevent 
damage to the screening. The one or two doors of the trap are 
left open but are closed just before daybreak. Many thousand 
anopheline mosquitoes have been taken daily inside such traps. 
It is wise to have an overhanging roof to protect the walls of 
the trap from rain, and to darken it somewhat so that the 
insects will not be too eager to escape at daybreak. 

Smaller traps, baited with a man sleeping inside but under a 
mosquito net, have also given good results. In one case we 
used a screened room in a nipa house, closing the windows at 

The malaria-carrying A. minimus var. ftavirostris apparently 
will take cow or carabao blood about as readily as it will take 
human blood, so that animal-baited traps are suitable even for 
this species. 

Baited traps located in the center of a control zone will serve 
as useful indicators to measure the success or failure of control 
projects. If the malaria-carrying mosquito is taken in the trap 
in considerable numbers the control project is not succeeding. 

If cattle were kept in sheds in the Philippines, such cattle 
sheds would doubtless be suitable places for catching adults, but 
we have rarely seen any sort of cattle shed in use locally. The 
usual shelters, if they exist at all, consist merely of a roof sup- 
ported by posts, with no side walls. Strong-material houses 
attractive as daytime-resting places for mosquitoes are also the 
exception rather than the rule in the provinces. 

Sometimes jungle species of Anopheles can be taken by sitting 
quietly in the dark in a suitable place, collecting those which 
attempt to feed. This is a dangerous procedure in malarious 

We have not had much success using collecting nets. Ap- 
parently the local species do not rest to any extent among 
grasses or shrubbery in the daytime. Occasionally it is possible 
to net considerable numbers of anophelines as they swarm beside 
a breeding place at dusk. 

We have not been successful with any of the usual plain box 
traps reported in the literature. Only our earth-lined box trap 
mentioned above has given us good results. 


Russell and Baisas: Handling Mosquitoes 


It requires a little experience to detect adult mosquitoes as 
they rest in their natural sheltering places. One must look 
closely and carefully, but it is not long before practice makes 
it easy not only to see the insects but even to differentiate 
Anopheles from Cidex and sometimes to identify immediately 
certain species. 


Whatever collecting tube is used it is necessary to have a box 
or a bag for transporting the insects to the laboratory. They 
must be kept out of the direct sun and should be handled as 
gently as possible. Various types of field transporting boxes 
have been used, and one may be 
easily devised to meet the needs. 
Wood is better than tin for a 
field box as it will not conduct 
heat so well. 

We have had considerable suc- 
cess shipping living mosquitoes 
by mail in individual glass vials, 
packed carefully in a mailing 
container. The vials we have 
found best measure 8 by 1.5 
centimeters. A hole is punched 
in the cork and a moist cotton 
plug is put in the hole. A strip 
of blotting paper is wedged in- 
side along the length of the vial 
and a mat of the same substance 
is placed at the bottom. We 
have had a high percentage of 
success by this method, up to 
three days in transit by train 
and bus, with outside shade tem- 
peratures up to 37° C. (98.2"* 
F.). As a matter of conven- 
ience the mosquitoes were 
caught and shipped in the same 
vials. After receipt in the laboratory the mosquitoes, still in 
their individual vials, if put in an ice box, will in many cases 
remain alive for a week (text fig. 5). 

We have had some success shipping live mosquitoes in a box 
patterned after Barraud's, but we have come to the conclu- 

Fig. 6- Vial for shipping living adults. 

276 The Philippine Jourwal of Science 1935 

sion, after comparative tests, that the individual vial method 
is a better one. 

Incidentally, if the mosquitoes being transported are from 
malarious regions, very great care must be exercised in opening 
the container which may contain deadly mosquitoes. It has been 
our practice to use a small screened cage having a cloth-sleeve 
entrance, so that both hands could be put inside the cage and 
any manipulation procedures carried on without danger of the 
mosquito escaping into the room. Of course, in many cases it 
will suffice to kill the insect in its vial by applying chloroform 
to the cotton plug in the cork, as soon as it is received in the 

The chief advantage in shipping mosquitoes alive is that they 
may be dissected in a central laboratory. It also facilitates 
studies of the ova. 

If it is desired to ship dead mosquitoes they may easily be 
killed with chloroform vapor. Sufficient exposure should be 
allowed so that the insects are actually killed and not merely 
stupefied. They are then put in pill boxes, one collection to a 
box, unless there are so many that undue crowding would result. 
Ordinary cardboard pill boxes may be used. A still better box 
is the small 0.25- or 0.5-ounce tin container such as is used 
to gather samples of faeces for diagnosis, or to hold various 

A little naphthalene, wrapped in a layer of cotton, is put in 
the bottom of the container. Lens paper is better. The insects 
are placed on top of the lens paper, or cotton, and are covered 
by another layer of the same material, and the tin or pill box is 
capped with its cover. Cotton expands and tends to put undue 
pressure on the specimens, so that if used it should be sparingly. 

A minute drop of creosote may be used instead of naphthalene 
as a deterrent against ants and other insects, which quickly 
destroy unprotected specimens in the Tropics. 


We have discussed the care of larvae in the laboratory up to 
the emergence of the adult insect. The adults that have emerged 
in the pupal vials may be transferred to lantern chimneys, 
which serve well as cages. The smaller end is covered with 
gauze held in place by adhesive tape. The lower, larger end 
rests in a Petri dish, the bottom of which is covered with filter 
paper. Raisins, boiled with sugar, seem to make acceptable 


Rtissell and Baisas: Handling Mosquitoes 


food for the insects. One boiled raisin put on top of the gauze 
will remain succulent for a considerable time. Not more than 
25 to 30 mosquitoes should be kept in one cage at a time. It 

Pig. 6. Lantern-chimney cage with manipulating tin. Dotted lines show position of base 
of chimney when adults are being inserted or withdrawn. 

is important to protect the cages from ants which can make 
serious inroiads during a single night, destroying and carrying 
off the adults. 

It is not necessary to use a lantern chimney. A small screened 
cage of any type can be used, provided food and water are 

Fig. 7. Boyd's manipulating forceps for removing adults from cage. 

supplied. The advantage of a lantern chimney is the ease of 
handling it. 

Boyd (22) has devised a test tube on forceps, which facilitates 
the transfer of mosquitoes from cage to cage (text fig. 7). If 

290499 i 

278 The Philippine Journal of Science 1935 

this mounted tube is used, the bottom of the chimney is covered 
with two pieces of rubber dam, or bathing cap. Each piece has 
a small slit in the middle and the two pieces are put together 
so that the slits are at right angles. This forms a valve opening 
through which the tube can be pushed to accomplish the removal 
or introduction of an insect. When the tube is withdrawn the 
valvelike slits close, preventing escape of mosquitoes from the 

Two flat pieces of tin or of cardboard, each with a hole near 
one end, may be of use in transferring insects from one chimney 
to another. They should be large enough, so that when part 
of the tin or cardboard is completely covering the base of the 
chimney, the end with the hole projects beyond. The gauze 
covering of the chimney is loosened and the chimney placed on 
the unperforated end of the piece of tin. The gauze can then 
be pulled away, the tin blocking the escape of the insects. The 
same procedure is followed for the other chimney. Then the 
two chimneys are placed end to end with the two pieces of tin 
intervening. By sliding these pieces of metal, the holes can be 
brought into apposition and the insects driven from one cage to 
the other. Such tins can also be used in transferring insects 
from a vial to a chimney, the vial being pushed through the 
hole (which should have a diameter only slightly larger than that 
of the vial). To replace the gauze, the chimney with its tem- 
porary tin bottom cover is placed on the flattened gauze. Then 
the tin is pulled away and the gauze fastened in place as before. 
With a little practice one becomes adept at manipulating the 
living insects. 

Aedes mosquitoes will mate in lantern chimneys, and it is 
relatively easy to propagate a colony in the laboratory. Culex 
mosquitoes require somewhat larger cages and can be propagated 
without much difficulty in cages measuring 1 by 0.5 by 0.5 
meter, but it is more difficult to propagate anophelines in cap- 
tivity. The most successful experiments have been made by 
Boyd. (23) Occasionally in the Tropics freshly hatched anoph- 
elines will mate in a small cage, and viable eggs will result; 
but this is exceptional. Apparently a very large cage with 
carefully controlled water and food supplies is required. The 
cage should be large enough to accommodate a carabao or even 
several animals and some nearly natural breeding places, such 
as a section of a stream, or a small pool. We have not had 
success in colonizing anophelines in captivity in the Philippines 
in a cage measuring about 6 by 6 by 10 meters. 


Russell and Baisas: Handling Mosquitoes 


To obtain eggs for study, fully gravid females should be 
selected and placed in a small cage, such as an oil cylinder glass 
or a lantern chimney, over water. Christophers (5) suggests 
floating in the water a large but thin paraffin-coated cork ring 
slightly smaller in circumference than the water container and 
held up by the meniscus. This greatly facilitates the micro- 
scopical examination of the ova in situ. A piece of paper or a 
stick in the cage will give the mosquito a foothold. (See also 
above under the discussion of ova.) 


FIG. 8) 

There are many ways of preserving and mounting adult anoph- 
elines. The following account closely follows Boyd. (4) 

Unmounted material can be 
kept either in the original pill 
boxes or transferred to test 
tubes. The former should be 
stowed in tight tin boxes, in 
which is kept a small dish of 
naphthalene and phenol to guard 
against insects and molds. If 
test tubes are used, they are 
prepared by pouring about 1 
cubic centimeter of melted naph- 
thalene into the bottom. This is 
covered by a small pledget of 
cotton. A layer of five or six 
mosquitoes is loosely arranged 
on this cotton, then another 
pledget of cotton is added, care being taken not to crush, but 
only to secure, the first mosquitoes. Alternate layers of mos- 
quitoes and cotton are inserted until the tube is filled. 

Specimens which have been stored and are dry may be pinned 
for mounting if they are first softened. Several layers of mois- 
tened filter paper are placed in the bottom of a dish having a 
tightly fitting cover. The dried mosquitoes are spread on this 
paper and the cover replaced. After a few hours the insects 
can be mounted with little danger of breaking the appendages. 

Pinned mounts are best made with freshly killed anophelines. 
When the mosquitoes are first exposed to chloroform vapor, the 
wings temporarily expand to a position at right angles to the 
body. MacGregor recommends that they be pinned immediately 

Fig. 8. One method of mounting an adult 
mosquito inside a glass vial. 

280 The Philippine Journal of Science 1935 

upon becoming stupefied, thereafter being returned to the chloro- 
form vapor until killed. Very fine pins 0.5 inch long, known 
by the German name "Minutenadeln," are best for insertion in 
the mosquito. These are without heads. Two methods of pin- 
ning may be followed. 

(1) Expose the ventral surface of the specimen, grasp a pin with the 
pinning forceps, and insert the pin point into the ventral thoracic waU 
between the legs, gently pushing until the point just emerges from the 
dorsum of the thorax. With the pin still grasped by forceps, its base 
is forced into and put through a small oblong piece of pith or cork near 
one end of the latter. A No. 2 full-length mounting pin is forced through 
the opposite end of the piece of pith or cork, until the latter is fixed at 
about the middle of the pin. The mounted specimens are best preserved 
in small vials, a separate vial being devoted to each mounted specimen. 
The heavy pin of the mount is forced into the inner surface of the stopper, 
in a slightly eccentric location. These vials are of clear, thin glass, with 
no constriction at the neck. A good size measures two inches in length 
by one inch in diameter. The label of the specimen may be transfixed by 
the mounting pin before its insertion into the stopper, or pasted on the 
outside of the vial. The stoppers may be horizontally perforated about 
3 mm, above the bottom, by a hole about three-eighths of an inch in dia- 
meter into which is poured melted naphthalene and phenol, which will 
quickly harden. 

(2) Using slightly heavier and full-length pins from which the head 
has been snipped, the insect may be pinned by thrusting the point trans- 
versely through the thorax, then thrusting the cut end of the pin into 
the vial stopper. By this method the thoracic scales receive less damage. 
The vials are conveniently stored in pasteboard boxes divided into com- 
partments, each compartment holding one vial. 

Less valuable specimens may be pinned in ordinary insect-proof boxes, 
in which, however, they cannot be as readily examined. 

Pinned specimens are not adapted to shipment through the mails. For 
this purpose specimens are best sent confined between layers of cotton in 
pasteboard pill boxes, a number of which can be shipped in a cylindrical 
mailing carton. 

A good deal of trouble has arisen in the Tropics from oxidiz- 
ing pins. The ordinary pins rust quickly. Silvered pins cor- 
rode in time and spoil the specimen. Nickel pins are best, 
but even these have not always proven satisfactory. It is now 
possible to buy stainless steel pins, but we have not yet tried 

In our experience we have had good results putting no pin at 
all through the insect. We fasten the specimens by a very small 
drop of shellac to the tip of a small wedge of cardboard. This 
small piece of cardboard is then fastened to the cork with a No. 
2 pin. The shellac is placed on the tip of the cardboard 
wedge, which is then carefully placed in contact with the 

56,3 Russell and Baisas: Handling Mosquitoes 281 

right side of the mosquito, which lies on its left side. To 
insure good adhesion the left side of the insect may then be 
pressed gently and carefully with a dissecting needle. This 
procedure, with a little practice, does no damage to important 
scales or bristles. It is possible to buy punches for cutting the 
cardboard wedges, but ordinary scissors may be used if neces- 
sary. Fairly long wedges are desirable when the insects are 
to be examined under high magnification. Practice is required 
in order to avoid damage to scales, wings, and legs, but when 
perfected the method gives excellent exposure for examination 
and the specimen does not deteriorate so rapidly as when pinned. 


We have prepared and distributed a wall-chart key to the adult 
anophelines of the Philippines and have also presented a paper 
on the same subject. (3) The latter contains full descriptions of 
each species and also a discussion of the characters used for 

A lens is required for the study of the specimens. A binocular 
entomological microscope is best, but an ordinary low-power 
monocular microscope can be used. In fact a strong doublet 
hand lens may be all that is required, although it will not give 
such good results as a microscope. 

The examination can be made in reflected daylight or under 
artificial light. We have had good results using a focusing 
flashlight with three or four dry cells. This gives a powerful 
light which brings the bristles and scales into good relief. The 
light can be easily directed at any angle and does not give off 
as much heat as the ordinary microscope lamps (Plate 6, fig. 1). 

With the specimen in good light under the microscope, the 
examiner must first be sure that it is a mosquito, and then, 
having decided that it is or is not an anopheline, note the sex. 
These three initial diagnoses are fundamental, and the following 
practical notes may be helpful: 


The following indicates that the place of mosquitoes among 
the insects is as follows: Phylum Insecta; order Diptera; 
suborder Nematocera; family Culicidse; subfamily 1. Dixinae; 
subfamily 2. Chaoborinse ; subfamily 3. Culicinas (the true 
mosquitoes) . 

The word "diptera" is from the Greek meaning **with two 
wings." Therefore, the first criterion is that the specimen has 
two wings (and six legs) . 

282 The Philippine Journal of Science 1935 

The ward "nematocera" is from the Greek meaning, loosely, 
"threadlike antenna." The Nematocera are small midgelike flies, 
with well-differentiated head, thorax, and abdomen, the last 
clearly segmented. The head bears two threadlike antennae, 
which have from six to sixteen segments and are usually covered 
with fine, long hairs. 

Next, to separate the Culicidse from the other families in the 
Nematocera we note that all Culicidae have a proboscis, long 
or short, and (except the Dixinse) all have a dense fringe of 
scales which project backward along the posterior margin of 
the wings. Moreover, the delicate membranous wings always 
carry a complete venation as shown in text fig. 3 of our key. (3) 
If a given specimen does not have the posterior fringe of wing 
scales and the typical venation, it is not a mosquito. 

Finally, to separate the subfamily Culicinse, or true mosquitoes, 
from the other Culicidse, note that all Culicinae, or mosquitoes, 
have a long proboscis, which about equals the combined length 
of the head and the thorax. 


The Culicinae are divided into three tribes; namely, Anophe- 
lini, Megarhinini, and Culicini. Edwards (24) gives the follow- 
ing key to these tribes : 

Key to the tribes of Culicinse. 

1. Abdomen without scales, or at least with the sternites largely bare. 

Abdomen with both tergites and sternites completely clothed with scales. 


2. Proboscis rigid, outer half more slender and bent backwards. 


Probscis more flexible, of uniform thickness (unless swollen at tip), 

outer half not bent Culicini. 

In our key to local adult anophelines we have mentioned 
other characters that may be of assistance in separating the 
tribe Anophelini. 


Most male mosquitoes can be recognized at once by their 
''whiskered" appearance. Their antennae have many long hairs, 
which give a distinctly plumed appearance easily noted at a 
glance. The female antennae are sparsely ornamented. There 
are a few mosquitoes (none of the Philippine Anopheles) in 
which the male and the female antennae are similar. The ques- 


Russell and Balsas: Handling Mosquitoes 


tion can always be settled by examination of the terminalia or 
the tip of the abdomen under a strong hand lens. 

The male terminalia have a somewhat broadened appearance. 
There are two conspicuous hooklike claspers which cannot be 
mistaken. The female genitalia are not conspicuous. The tip 
of the abdomen is more pointed and there are seei^ two small 
papillae, called cerci, which project either vertically or longitu- 
dinally (text fig. 9). 

Fig. 9. 

Tip of abdomen of anopheline adults, male on right, female on left, showing dis- 
tinctive characters. 

Once the specimen in question is known to be an anopheline 
mosquito, it may be examined systematically with our key (3) at 
hand. When by means of the key a probable diagnosis of species 
is arrived at, the descriptive text referring to that species may 
be consulted to check in detail the various characters. When the 
drawings given in the corresponding plate are consulted, in 
most cases there will remain no doubt as to the identification of 
the specimen. 

In all doubtful cases the specimen should be sent to the Health 
Service or to the Bureau of Science for further study, together 
with as many details about habitat as possible. Dried mounted 


The Philippine Journal of Science 




►si 03 





P. •_^ 



Russell and Balsas: HaTidling Mosquitoes 


p g 

- 5 

-& 2 



S be 

I 6 

s - 

^ w 

fi •• &» 

* ^^ ^ 


.. ^ c 
. <y ti 

•g 2 5 
"> £ . 

TJ g 

^'^ § 
g .. ^ 

g 42.C 

be ^ 

i* '^ tC 

« " s 
8 g g 

d 0, , 

286 The Philippine Journal of Science 1935 

mosquitoes usually do not have their abdomen in the right posi- 
tion for examination. In such cases the entire abdomen is cut 
off and heated, first in a 10 per cent potassium hydroxide solu- 
tion, and subsequently in distilled water. This process renders 
the organ in excellent condition for examination. The hairs 
and scales are not removed, the whole abdomen is transparent, 
and even the color of the scales is retained. For future reference 
such specimens may be preserved in formalin-borax solution or 
weak alcohol, and kept in a properly labeled small vial. 


Gut, — There are many ways of dissecting out the gut of a 
mosquito, and each technician has his own ideas as to which 
is best. The following description is based on Boyd's (4) direc- 

FlCL 12. Position of microscope, hands, slide, needles, and insect during dissection. 

The specimens to be dissected are killed one at a time as 
needed, with the use of chloroform vapor. It is much better 
not to attempt to use those that have a gut distended with 
blood, but to dissect only those that have fully digested their 
last blood meal. Careful note should be taken of the species, 
as this is very important. In the Philippines, so far, only A. 
minimus var. flavirostris and A. maculatus have been found 
infected in nature. While the insect is held with forceps, which 
grasp the palps and proboscis, the wings and legs are cut off 
close to the body. 

The mosquito is placed on a glass slide in a drop of physiol- 
ogical saline solution, which has been deeply tinted with methyl- 
ene blue, and arranged with dissecting needless so that the dor- 
sum of the insect is away from the dissector and the tip of the 

56,3 Russell and Baisas: Handling Mosquitoes 287 

abdomen at the right. Under a low-power lens the insect is 
held firmly by means of the straight needle thrust into the thorax 
near the abdomen. With the curved needle cuts are made into 
the seventh or eighth abdominal segment, both dorsally and 
ventrally. Then the last segment of the abdomen is transfixed 
with the tip of the curved needle, and the viscera attached to 
this abdominal tip are drawn out. The f oregut usually breaks, 
and it is easy to draw out the viscera into the saline solution. 
(Some dissectors prefer to cut the abdomen away from the 
thorax before dissecting out the gut.) 

The body of the mosquito is put to one side and the gut 
cleaned by gently cutting off the Malpighian tubes and other 
structures posterior to them. The clean gut or "stomach'' is 
transferred by means of the needle to a drop of untinted saline 
solution on a clean slide and protected by a cover slip. The 
slide is gently heated over an alcohol lamp so that the heat will 
expand the gases inside the gut and so extend the walls, making 
them thinner. 

The preparation should next be searched under high power for 
oocysts. These cysts vary in size from about 6 to 50 microns. 
They are approximately spherical and appear a little more 
opaque than the gut cells. Because they are attached to the 
stomach wall they are usually distinct from yeast cells or similar 
bodies contained within the cavity of the stomach. One of the 
most useful criteria for the diagnosis of an oocyst is the clump 
or chain of brown or black pigment granules which are always 
present in all but the very large oocysts. These very large 
cysts are mature and striated with innumerable sporozoites. 
Such large cysts may be ruptured by slight pressure and the 
sporozoites, as they escape, are recognized by their morphology, 
as noted below. A small cyst should never be diagnosed as a 
malarial oocyst unless it is definitely pigmented and is seen to 
be attached to the stomach wall. Sometimes, if the stomach 
is rolled a little, by pressing a needle on the cover slip, a doubt- 
ful body may be seen not to be attached to the wall. 

Fat droplets and immature ova on or under the stomach, 
yeast cells in the stomach cavity, and occasionally other bodies 
may simulate oocysts; but if the cyst is definitely attached to 
the stomach wall, and if it certainly contains either pigment 
granules or recognizable sporozoites, there can be little confusion. 
Small oocysts up to 7 or 8 [x have no definite cyst wall. A single 
stomach may show from one to more than one hundred oocysts 
(Plate 7). 

288 3"/ie Philippine Journal of Science 1935 

Salivary glands, — What remains of the mosquito from which 
the gut was dissected should be turned so that the head and 
thorax project towards the right. The specimen is held in place 
by transfixing the posterior part of the thorax with the straight 
needle held in the left hand and the head carefully cut off with 
one clean, quick cut with an edged needle. The dull side of 
a needle held in the right hand is pressed gently on the middle 
of the thorax, forcing a small amount of tissue out at the neck 
end. Under a lens this tissue is cut off and gently teased apart, 
the examiner being alert to recognize the two blue-stained and 
trilobed salivary glands, which will show up clearly among the 
unstained muscle fibers. Then the glands are transferred to a 
drop of untinted saline solution on a clean slide, covered with 
a cover glass, and examined for sporozoites (Plate 8, figs. 1 
and 2). 

The thin fusiform sporozoites are usually easily recognized. 
They are either straight or slightly curved and average 12 to 
14 [L in length. When alive they have a slow, sinuous move- 
ment. They must be distinguished from such artifacts as rod- 
shaped bacilli or crystals. It is always a good plan to stain 
them with Wright's or Giemsa's stain. Each sporozoite will 
show blue or violet cytoplasm and red chromatin. To stain, 
the cover is carefully lifted and inverted on a dry slide on 
a tiny drop of balsam. The balsam holds the cover slip in 
position while the preparation is dried, fixed, and stained. 
For description and methods of preserving these glands, see 

It is quite possible to dissect out the glands quickly and easily 
simply by cutting off the head, as the mosquito lies on its side 
in a drop of saline solution, and pressing gently on the pleura 
of the thorax, expressing the glands, often well isolated. It is 
not even necessary to remove the legs or wings, after a little 
practice. We have had good results by this simple technic when 
we have had many specimens to examine in a short time. But 
much better preparations are secured by the method outlined 

Salivary-gland dissections may be dried on the slide, fixed, 
and subsequently stained with Wright's or Giemsa's stain, and 
sent to a central laboratory for confirmation of diagnosis. 

The dissection of salivary glands of Culex mosquitoes is more 
difficult and requires additional practice. 

Terminalia. — The dissection and preparation of the male 
terminalia are not easy and require a great deal of practice. 

56,3 Russell and Baisas: Handling Mosquitoes 289 

After the terminal segment has been cut off it is heated in a 10 
per cent solution of potassium hydroxide. Care is taken not to 
heat it beyond the first steaming point, the correct temperature 
being about 60° C, such that the finger can just bear to touch 
the dish. Overheating makes such delicate parts as the leaflets 
of the phallosome very transparent so that it cannot be deter- 
mined whether the thinner edges are serrated or not. 

The phallosome is severed from its basal connections with the 
coxites, the theca is cut short, after which the apical portion, 
together with the leaflets, is transferred to a clean slide, with 
a tiny drop of Gater's fluid. The material is carefully immersed 
in the fluid and watched through a binocular while a cover slip 
is put in place. With a stout needle the cover glass is pressed 
down firmly when the phallosome is in the right position. The 
pressure helps to spread the leaflets nicely on either side without 
the need of cutting the theca into two longitudinal halves. The 
mounting fluid should be so little as to spread barely to the 
edges of the cover slip when pressed. Additional fluid may be 
added to make up the shrinkage after drying. Care should be 
exercised to use perfectly flat cover slips, as those which are 
a little convex spoil the best dissections by not staying in posi- 
tion when pressure is removed. 

The isolation of the harpago is best done by pressing the 
outer side of the coxite with a needle directed inwardly at an 
angle of about 45°. This tilts the coxite a little and exposes the 
harpago to the best advantage for dissection. 


The technic of handling mosquitoes requires patience and 
practice, but it is not, after all, very difficult. It is essential 
that one have available a small library. If funds are limited, 
we should suggest as a minimum the following list: 

Handbooks: Boyd's Malariology(4) and MacGregor's Mos- 
quito Surveys. (27) Reprints: King,(i2,a,b,c) Manalang,(l0,25, 
26) and Russell and Baisas. (1,2, 3) Wall charts: Key to Larvse 
of Philippine Anopheles and Key to Adult Philippine Anopheles, 
published by Malaria Investigations and obtainable through the 
Bureau of Health. 

The investigation of Philippine mosquitoes is a matter of im- 
portance, and so little has been done that it is a fruitful field 
for research. 

290 The Philippine Journal of Science 1935 


1. Russell, P. F., and F. E. Baisas. Habitats of Philippine Anopheles 

larvae. Philip. Journ. Sci. 55 (1934) 297-306. 

2. Russell, P. F., and F. E. Baisas. A practical illustrated key to lar- 

va3 of Philippine Anopheles. Philip. Journ. Sci. 55 (1934) 307-336. 

3. Russell, P. F., and F» E. Baisas, A practical illustrated key to Phil- 

ippine Anopheles adults. To be published in Philip. Journ. Sci. 

4. Boyd, M. F. An Introduction to Malariology. Harvard Univ. Press, 

Cambridge, Mass. (1930) 437 pages. 

5. Christophers, S. R. Family Culicidae, Tribe Anophelini. Vol. IV. 

Diptera in the Fauna of British India. Taylor and Francis, Lon- 
don (1933) 371 pages. 

6. Gater, B. a. R. Aids to identification of anopheline larvas in Ma- 

laya. Govt. Print. Office, Singapore (1934) 160 pages, 

7. Hackett, L. W. The present status of our knowledge of the sub-spe- 

cies of Anopheles maculipennis. Trans. Roy. Soc. Trop. Med. & Hyg. 
28 (1934) 109-140. 

8. Russell, P. F., and D. Santiago. Anopheles minimus larvas from 

wells in Laguna Province, Philippine Islands. Philip. Journ. Sci. 
49 (1932) 219-223. 

9. Russell, P. F., and D. Santiago. Flight range of Anopheles in the 

Philippines. Second experiment with stained mosquitoes. Am. Journ. 
Trop. Med. 14 (1934) 407-424. 

10. Manalang, C. Morphology and classification of the Philippine variety 

of Anopheles aconitus Donitz, 1902, and Anopheles minimus Theobald, 
1901. Philip. Journ. Sci. 43 (1930) 247-261. 

11. Baisas, F. E. The barbirostris-hyrcanus group of the Philippine Ano- 

pheles. Philip. Journ. Sci. 44 (1931) 425-448. 

12. King, W. V. (a) The Philippine varieties of Anopheles gigas and 

Anopheles lindesayi. Philip. Journ. Sci. 46 (1931) 751-757. 

(b) The Philippine Anopheles of the rossi-ludlowi group. Philip. 
Journ. Sci. 47 (1932) 305-342. 

(c) Three Philippine Anopheles of the funestus-minimus subgroup. 
Philip. Journ. Sci. 48 (1932) 485-523. 

13. Russell, P. F. Identification of the larvae of the three common ano- 

pheline mosquitoes of the Southern United States. Am. Journ. Hyg. 
5 (1925) 149-174. 

14. Barber, M. A., et al. Distribution of the commoner anophelines and 

the distribution of malaria in the Philippine Islands. Philip. Journ. 
Sci. 10 (1915) 177-247. 

15. Holt, R. L., and P. F. Russell. Malaria and Anopheles reconnais- 

sance in the Philippines. Philip. Journ. Sci. 49 (1932) 305-371. 

16. Russell, P. F. Malaria and Anopheles reconnaissance in the Philip- 

pines, II. Philip. Journ. Sci. 54 (1984) 43-59. 

17. Boyd, M. F., and H. Foot,. The alimentation of anopheline larvae and 

its relation to their distribution in nature. Journ. Prev. Med. 2 
(1928) 219-242. 

18. Russell, P. F. Daytime resting places of Anopheles mosquitoes in the 

Philippines. First report. Philip. Journ. Sci. 46 (1931) 639-649. 

56,3 Russell and Baisas: Handling Mosquitoes 291 

19. Russell, P. F. Daytime resting places of Anopheles mosquitoes in 

the Philippines (Second report). Proc. Ent. Soc. Washington 34 
(1932) 129-138. 

20. Russell, P. F., and D. Santiago. Flight range of the funestus-mini- 

mus subgroup of Anopheles in the Philippines. First experiment with 
stained mosquitoes. Am. Journ. Trop. Med. 14 (1934) 139-157. 

21. Russell, P. F., and D. Santiago. An earth-lined trap for anopheline 

mosquitoes. Proc. Ent. Soc. Washington 36 (1934) 1-15. 

22. Boyd, M. F. Methods for the manipulation and conservation of ano- 

pheline imagines. Am. Journ. Hyg. 16 (1932) 839-844. 

23. Boyd, M. F., and T. L. Cain, Jr. On large scale rearing of Anopheles 

quadrimaculatus in captivity. Am. Journ. Hyg. 16 (1932) 832-835. 

24. Edwards, F. W. Diptera. Family Culicidae. Genera Insectorum (P. 

Wytsman) Desmet-Verteneuil, Bruxelles (1932) 258 pages. 

25. Manalang, C. Notes on malaria transmission. Philip. Journ. Sci. 37 

(1928) 123-131. 

26. Manalang, C. The preservation of malarial oocysts and sporozoites. 

Philip. Journ. Sci. 42 (1930) 481-487. 

27. MacGregor, M. E. Mosquito Surveys. Bailliere, Tindall and Cox, 

London (1927) 282 pages. 


Plate 1 

Fig. 1. Collecting larvae, using dipper on the end of a rattan cane, making 
wide sweeps of surface water. 

2. Collecting larvae, dipper being pushed against a bank so that water 

flows into it from among the roots of the vegetation. 

3. Carrying case for larva-collecting bottles. 

Plate 2 
Fig. 1. Larva-collecting outfit. Three types of dipper, strong rattan cane, 
collecting pipettes, half-pint collecting bottles, knapsack carrying 
case, and rubber boots. 
2. Rearing outfit for larvae. Collecting bottle, enamel pans, glassware, 
pupal tubes, and pipettes. 

Plate 3 
Fig. 1. Collecting adults. Showing a typical Anopheles minimus var. fla- 
virostris breeding place and a collector using glass and rubber 
tubing, taking adults from a daytime shelter in small caves formed 
by an undercut stream bank. 

2. Collecting adults with a glass-and-rubber, suction, catching tube. 

Natural daytime shelter. 

3. Catching adults with a small vial. Natural daytime shelter. 

Plate 4 
Fig. 1. Anopheles adult trap, at Iwahig, suitable for carabao bait, distant 

2. Anopheles adult trap suitable for carabao bait, near view. 

3. Earth-lined Anopheles adult trap, with cover held up. 

Plate 5 
Fig. 1. Catching of adult Anopheles on carabaos at night. (Photograph 
taken in daytime.) 
2. Lantern chimney, tin cover for transferring mosquitoes, and wire 
cage with cloth-sleeve entrance. Adults that may be infected 
are transferred from vials to lantern-chimney cage inside this 
larger cage. 

Plate 6 
Fig. 1. Flashlight used as a source of illumination for the examination 
of adult mosquitoes. Method of placing a specimen for exam- 
ination and the type of binocular microscope suitable for this 
purpose. A simple stand to hold the flashlight can easily be 
2. Method of mounting an adult mosquito in a small drop of shellae^ 
as explained in the text. 

290499—5 293 

294 The Philippine Journal of Science 

Plate 7 

Numerous oocysts of malaria on the stomach wall of AnopheleB minimut 
var. flavirostris. 


Fig. 1. A mature oocyst under higher magnification, showing sporozoites. 

[From Manalang.(25)] 
2. Sporozoites from a ruptured salivary gland; X 1,000. Stained. 

[Prom Manalang.(25)] 
8. Sporozoites from a ruptured salivary gland, in saline solution; X 

1,000. Unstained. [From Manalang. (26) ] 


Fig. 1. Ctdex egg raft 

2. Two types of eggs of Anopheles minimua var. flaviroBtria, (Drawn 

by Andres Nono.) 
8. Special type of dipper. 

4. Larva shipping bottle. 

5. Vial for shipping living adults. 

6. Lantern-chimney cage with manipulating tin. Dotted lines show 

position of base of chimney when adults are being inserted or 

7. Boyd's manipulating forceps for removing adults from cage. 

8. One method of mounting an adult mosquito inside a glass vial. 

9. Tip of abdomen of anopheline adults, male on right, female on 

left, showing distinctive characters. 

10. Adult mosquito, showing, p, proboscis; pp, palps; a, antenna, 

c, clypeus; e, eye; h, head; M, haltere; th, thorax; ist, thoracic 
stigmas; ah, abdomen; ast, abdominal stigmata; w, wings; II, 
stumps of legs. (After Nuttall and Shipley.) 

11. Gross internal anatomy of mosquito, showing, pp, palp; md, man- 

dibles; mx, maxillae; he, buccal cavity; ph, pharynx; sd, salivary 
duct; w, oesophagus; sg, salivary glands; dr, dorsal reservoirs; 
vr, ventral reservoir; ca, caeca; mg\ midgut begins; mg\ mid- 
gut ends; st, stomach; mt, Malpighian tube; i, ileum; co, colon; 
T, rectum; a, anus. (After Nuttall and Shipley.) 

12. Position of microscope, hands, slide, needles, and insect during dis- 



! riiiLiP. .loriiN. Sri., 5r», No. 3 




i ■'' " -iii' 


ErssFi.L AN}» 1*a;sas: IIamh.ini: MiKstiiiToh's. | 

I P!Mi,ii*. J*TK\, S<1., .'iIk No. 



-4. I- > ^'/^ 

f S..:| 

€ ^:% 





I Philip. J<H'RN. S<i., SIJ, N<*, 3 

^^ • . -I - ? 

' ^%' ti§^ 




?>-".|^-' ■" 

^^ .:k- 


. -i«5«'!r?-: <jt;'^"' rir"=-f''^""4i----".-' " 

llf'HHKLL ANIi BaiSAH : H.\N!>l.lKr, MoSt|l!ItOKS. ] 

IPfllLll*. JoURN. SC'L, Mi, Nil. 3 


.;?^: ■ ■ 



Pniijp, JofiRx. S<"L. •'><>, No. S 


RlTSSKLL AN1> BaISAK : llANDIJNr, M«>St|!'rr<>KS. | 

[Pmi.iF. Joi'EN. S(i.. nf^ No. 3 



Russell and Bmaas: Hanhlint, Mosqi-itok.'^. | 

fPlllLIP. JlHJRN. Sci., S6. No. 3 



.-, ^;"iAf: 



" '*^^^^ 





By EusEBio Y. Garcia 

Of the Department of Parasitology, School of Hygiene and Public Health 
University of the Philippines, Manila 


In an experiment of this nature it is important to establish 
a safe and satisfactory criterion for determining the viability of 
the cysts. A review of the literature on the subject reveals that 
there is no standard criterion by which to determine whether 
a cyst is dead or living. Various authorities who have worked 
on the determination of the longevity of Entamoeba histolytica 
cysts in vitro have reported various results with widely varying 
significant differences, probably because different criteria were 
used in judging the viability of the cysts. 

Kuenen and Swellengrebel (1913), working on the hypothesis 
that dead cysts stain with a weak solution of eosin in water 
(1 : 1,000) and that living cysts do not, found that all the cysts 
in faeces begin to degenerate in from thirteen to twenty-nine 
days. The reliability of this result was later discounted by 
Yorke and Adams (1926), who gave as a proof of their conten- 
tion the nonexcystation of cysts that they believed to be living 
in the Locke-egg-serum medium of Boeck and Drbohlav (1925). 
Penfold, Woodcock, and Drew (1916), determining the viability 
of the cysts by their power to excyst in the presence of liquor 
pepticus, stated that although they were able to keep cysts in 
slow-running water for fifteen days, some certainly were alive 
at the end of this period. Wenyon and O'Connor (1917), em- 
ploying the eosin test for viability, claimed that cysts will sur- 
vive for over thirty days in water. Dobell (1919), cited by 
Yorke and Adams (1926), employing the eosin test supple- 
mented by morphological observations, found that the maximum 
viability of cysts in water is five weeks, and that they may be 
found dead at the end of a week in faeces. Boeck (1921), also 

* Read before the Third Philippine Science Convention, held in Manila 
February 28, 1985. 


296 The Philippine Journal of ScienAie 1935 

using the eosin test supplemented by morphological observations, 
stated that E. histolytica cysts were viable at the end of one 
hundred fifty-three days in water. This very high figure led 
some investigators to doubt the accuracy of such results, and in 
1932 Wight and Wight, after observing the viability of the 
cysts by actual culture, claimed that Boeck never proved that 
after this long period (153 days) the cysts that he believed sur- 
vived were capable of perpetuating the species. Long after the 
cysts are moribund or dead, the wall may retain impermeability 
for certain dyes. Sellards and Theiler (1924) modified the 
method of Kuenen and Swellengrebel for determining the viabil- 
ity of the cysts by infecting kittens by intrarectal injections of 
the cysts instead of per os, and to their disappointment they 
failed to produce infection even with the use of quite fresh 
material. Wight and Wight (1932) seemed able to prove that 
the conclusion of Yorke and Adams (1926), based on cultural 
methods, were entirely inadequate, and that Dobell's figure (37 
days plus) , based on the eosin test seemed to be nearer the truth. 
The same authors claimed that the majority of the dead cysts of 
Yorke and Adams on Locke-egg-serum medium were still viable 
when subcultures were made in the same medium. From the 
latter observation we may infer that cultural methods, no mat- 
ter how favorable the medium, are not a satisfactory criterion 
for determining the viability of cysts, because all cysts that have 
been exposed to the action of one chemical or drug, or any test 
solution or test medium, will not excyst at the same time, as not 
all of them have the same resistance. This inference seems to 
be borne out by the fact that the same authors (Wight and 
Wight, 1932) found that the preexcy sting forms in Locke-egg- 
serum medium are similar to dead cysts and vice versa. 

From these contradictory, not to say confusing, reports, it 
seems clear that the results of the different investigators are 
much at variance. In the writer's opinion these varying results 
are due to the lack of a standard criterion for judging the 
viability of the cysts. The writer is carrying on experiments 
along similar lines but using for determining the viability of 
cysts an entirely different criterion, based on the principle of 
cell necrobiosis, which, in the writer's opinion, is more reliable 
than the criteria used hitherto. Cystologists and physiologists 
have demonstrated that the nucleus as a whole is the governing 
body of the cell, and without it the cell is considered dead and 
beyond resuscitation. No one would deny that the cyst is in the 

56, 3 Garcia: Chlorinated Lime and Amoeba 297 

dormant and resistant state of a living cell which has met 
unfavorable conditions. That being the case, the nucleus is the 
most important structure of the cyst. If parts of the proto- 
plasm of a cell are lost, leaving only a very thin rim around the 
nucleus, as long as the nucleus remains intact the cell is still 
living, and able to regenerate and to perform its physiological 
functions. If the nucleus is lost, the whole cell is lost. In 
other words, a dead cyst may be defined as one whose nucleus 
or nuclei have been destroyed or have disappeared, and whose 
cytoplasm is shrunken, coarsely granular, and vacuolated. Cysts 
that do not answer to this definition are not considered dead. 

An opportunity to test the reliability and usefulness of a cri- 
terion for the determination of cyst viability presented itself 
when Dr. A. Ocampo, of the Philippine General Hospital, sup- 
plied the writer with faeces from a confirmed case of entamoebic 
dysentery. There could be no doubt that the patient was suf- 
fering from chronic amoebiasis. The cysts present in the fseces 
answered all the morphological descriptions of typical E. his- 
tolytica cysts; and, therefore, there seems to be no doubt that 
E. histolytica cysts were used in this study. 

In view of the fact that the results of Wenyon and O'Connor 
(1926), that chlorinated lime tabloids as used for water steriliza- 
tion had no action on E. histolytica cysts, seem unreliable, be- 
cause their findings were based on the use of eosin as a test 
for viability; and that the results of Yorke and Adams (1926) 
are subject to the same criticism, as proven by Wight and 
Wight (1932), the writer felt justified in making further ob- 
servations on the action of chlorinated lime on E. histolytica 
cysts. The writer was also influenced by his interest to deter- 
mine the lowest concentrations lethal to the cysts, because he 
believes this is important from the standpoint of water steriliza- 
tion by the use of chlorinated lime, which is now being used in 
large and small filtration plants and in the sterilization of swim- 
ming pools. 

In order to forestall errors in this experiment an entirely 
different line of approach was followed. The chemistry of hypo- 
chlorite of lime (CaOCl2), especially its rate of solubility in free 
chlorine water, was studied, and it was found that 1 gram of 
the powder will completely dissolve in forty-five minutes |n 100 
cubic centimeters of water without stirring, and in thirty min- 
utes with stirring. These findings approximate those of Bus- 
well (1923). According to Buswell CaOCl2 is not very soluble 


The Philippine Journal of Science 




. 8i<p fl f3 V 

§•2-11 S 





'%U99 jaj; 







•^aaa i9^ B 


'%u&a jtaj 


"2 2 ^ ^ I 

3 _, « *^ U 

i A A, Ai A A : 


66, S 

Garcia: Chlorinated Lime and Amceba 





Distorted and 
broken; four 
nude! disap* 
peared; chro- 
matold bodies 
d 1 s appeared; 
very much 
shrunken and 

d 5 - 

Sf !1 


o- * 2 « 

'3- Bali's SI:S g 


a S A 





The Philippine Journal of Stdence 








45 1" 

§ § 












'^UdO J3J 


S rt N ?fi Q» — ? 

^ S 


t, ^ 

i^ -O 2 Pi 
"S „ ■, _ . 

•:^uaD ja^ 



S ^ o S ^- S -g 


t> O 01 (8 ^ 

d «« T3 ** ? ^ .2 

-" ft a no « M > 

a s 

1 s 

.« .o c 

^ -M ft 

•:juao ja«i 


«Si^oft6«i)»:'2 43 

£3 *^ 

"3 - 

•« 8 g 


a '-^ ^ 

I c a I g p « § " 

*» M B ftft^'S ^Ij'd 

too fHM c3rja«!S«i3 


Garcia: Chlorinated Lime and Arrweha 


L o 

-^ ra *« S i 

; 3 ^ 2 § 

-s a •« 



s^ s e 

-5 3 § i: 


ja p. - 


O p « 0! ea .« 

2 i 

M ^ .S S « 

••» * * OS 

? » It" i 

i3 tt ft a tt) > 

— -d 


1 § 

^ -*^ 

^ ft (U > 0*0 










« S o ^ 

•-S s -e fl 

.- a> T3 

c a £ 


— w C 

;. M v. « -^ 

^ a 

ft ft ^ 

I ^ i5 ^ ft 

> tJ 


The Philippine Journal of Science 








o « 











'9(193 ia^ 




•3 e$ 


fl a § I g « o 


.4 w C 

: M m 

p. -. 

>» .S 

53 "S 

2 i (i. ^ ^ s -s 

? S ^ ''' I I fl I ? I 

I « a 1 1 I s i : 

g 2 ^ 


8 ^ 

^ ^ 

•%aaa 39^ 

'%osa ja j 



I 5 ^ S l-ft- •! ► •§ 5 ^ 

S6, 3 

Garcia: Chlorinated Lime and Ammha 







§ '^ ; 

t3 « j 

Two nuclei dis- 
torted and two 
broken; cyto- 
pli^m shrunk- 
en and slightly 
vacuo lated; 
chromatoid bo- 
dies fragment- 

Four nuclei dis- 
torted; karyo- 
8 o m e disap- 
peared; cyto- 
plasm shrunk- 
en and slightly 




Two nuclei dis- 
torted and two 
broken; cyto- 
plasm shrunk- 
en and slightly 
chromatoid bo- 
dies fragment- 

Four nuclei dis- 
torted; karyo- 
8 o m e disap- 
peared; cyto- 
plasm shrunk- 
en and slightly 

a s s 

S M 

Two nuclei dis- 
torted and two 
broken; cyto- 
plasm shrunk- 
en and slightly 
vacuo lated; 
chromatoid bo- 
dies fragment- 

One nucleus bro- 
ken and three 
distorted; cy- 
toplasm shrunk- 
en and slightly 
chromatoid bo- 
dies fragment- 

Four nuclei dis- 
torted; karyo- 
8 o m e disap- 
peared; cyto- 
plasm shrunk- 
en and slightly 

s s 


Two nuclei dis- 
torted and two 
broken; cyto- 
plasm shrunk- 
en and slightly 
vacuo lated; 
chromatoid bo- 
dies fragment- 

Four nuclei dis- 
torted; karyo- 
8 o m e disap- 
peared; cyto- 
plasm shrunk- 
en and slightly 



304 The Philippine Journal of Science 1935 

and is ordinarily fed as 1 or 2 per cent suspension. This is 
important to know, because it will enable us to predetermine 
the strength of the dilutions to be used to test the action of this 
chemical on the cysts. Judging from the results of Wenyon and 
O'Connor (1926) and Yorke and Adams (1926) it seems rea- 
sonable to suppose that this important fact had been ignored 
by these authors. 

Another important feature in this work was the use of nat- 
ural faeces instead of washed and concentrated suspensions, as 
done by Yorke and Adams (1926) . While the use of the latter is 
more ideal than the use of the former, the writer believes, as he 
points out in his discussion of the' results, that the presence of 
nitrogenous organic matter in the fseces is most necessary in the 
action of CaOCl2 on the cysts. Some chemical substances are 
formed in the interaction between calcium hypochlorite and ni- 
trogenous organic matter, and these have a synergistic killing 
effect on E. histolytica cysts. The presence of nitrogenous 
organic matter in the fseces is always certain, because bacterial 
decomposition of proteins takes place in the fseces and the bac- 
teria are capable of deamidizing these nitrogenous substances 
into free ammonia (NH3) and fatty acids. 


Sixteen-hour-old fseces of jelly like consistency were stirred 
thoroughly with a platinum loop 5 millimeters in diameter. One 
loopful of the fsecal emulsion was spread uniformly on a large 
slide, covered with a large cover glass, and sealed with paraffin. 
A total count of the cysts was made under a high-power lens after 
the manner of StolFs egg-counting technic ; that is, by the use of 
a square fenestra attached to the eye piece and a cliquer. 
Counts were made on several such preparations, and it was found 
that each loopful contained an average of fifty-six cysts. The 
types of cysts present were 96 per cent quadrinucleates and 4 
per cent binucleates (Plate 1, figs. 1 and 2) . 

After the average number of cysts per loopful was determined, 
1 per cent solution of calcium hypochlorite (CaOCl2) in dis- 
tilled water was prepared by shaking the preparation vigorously 
in a tightly stoppered flask until the CaOCl2 was completely dis- 
solved. From this original solution varying dilutions up to 
1 : 1,000,000 were made. Five cubic centimeters of each of these 
test solutions were placed in a tightly stoppered test tube and 
the series arranged in sequence in a test-tube rack. Four loop- 


Garcia: Chlorinated Lime and Amoeba 


fuls of the faeces were inoculated to each test tube containing 
the test solution. By means of a capillary pipette two drops 
were taken from the bottom of each test solution after thirty 
minutes and spurted on a slide, covered with a cover glass, and 
examined under an oil-immersion objective. A careful study 
was made of the internal organization of each cyst that came 
under observation, and the ratio between living cysts and the 
dead ones determined by use of the aforementioned criterion. 
This procedure was repeated in another series, in which washed 
and concentrated suspensions of E, histolytica cysts were used 
instead of faeces containing cysts. The results are shown in 
Tables 1 and 2. 

Table 2. — Percentages of different types of washed Entamoeba hystolytica 
cysts that were observed in different dilutions of hypochlorite of lime. 

Hypochlorite of 

lime dilutions 

(5 cc). 

Typea of Entamoeba histolytica cysts. 

Control. Concentrated 
and washed E. histo- 
lytica cysts with 96 
per cent = 4 nuclei, 
4 per cent == 2 nuclei. 







1 hr. 





1 :100 

1 : 1,000 

1 : 10,000 

1 : 100,000 

1 : 1,000,000 




















No change. 
























Table 1 shows that while striking changes could be noted 
in the structure of the cysts exposed to different tests solutions, 
no changes at all occurred in those in the control tubes. The 
highest dilution of calcium hypochlorite (CaOCl2) that will kill 
E. histolytica cysts after one hour fifteen minutes exposure is 
1 : 100,000. This concentration, when reduced to parts per mil- 
lion, is equal to 10 parts per million; or, when expressed in 
chlorine, it is equal to 3.5 parts per million, because the available 

306 The Philippine Journal of Science issb 

chlorine present in the solution is 35 per cent. Aside from these 
striking changes, several types of degenerative changes in the 
cysts were found, not only in single test solutions but in all of 
them at different exposures. The results indicate beyond a doubt 
that calcium hypochlorite (CaOCl2) really exerts a deleterious 
influence on E. histolytica cysts. 

The efficiency of CaOC^ in killing E, histolytica cysts noted 
in this work is attributed by the writer to the fact that cysts 
were inoculated immediately after this chemical has been com- 
pletely dissolved, allowing a maximum ionic action of the solu- 
tion on the exposed cysts. It is presumed that either the hypo- 
chlorous (HCIO) or chlorine (CI) ions may penetrate the cyst 
wall and there act on the protoplasm of the cyst by oxidation. 
The chlorine (CI2) excess liberated by CaOCl2 that does not 
act on cysts may by its addition and substitution reactions act 
on the decomposition products of protein, which are mostly 
nitrogenous organic compounds present in the faeces. Dakin 
et al. (1916) found that the combination of CI2 with nitrogenous 
organic matter forma strong antiseptic chloramines, which have 
stronger bactericidal action than chlorine, due to their greater 
solubility and penetrating power. They do not decompose spon- 
taneously as does HCIO and, therefore, their entire strength is 
maintained for a longer time. In this work greater action of 
CaOCl2 was expected because the preparation that contained 
the cysts also contained organic matter. 

By repeating the experiment of Yorke and Adams (1926) 
simultaneously with the first experiment, the writer found that 
only a 1 : 100 dilution of CaOCl2 can kill E. histolytica cysts af- 
ter one hour fifteen minutes exposure (Table 2) . This result 
partly confirmed their conclusion in 1926 that only high concen- 
trations of CaOCl2 can kill E. histolytica cysts. It seems, how- 
ever, that in the 1 : 100 dilution of CaOCl2 the action of the 
chemical is not wholly due to CI2, but partly to traces of chlo- 
ramines formed from the small quantity of nitrogenous organic 
matter which was left with the cysts after several washings and 
centrifugations. It is natural to expect, therefore, that in a con- 
centrated faecal emulsion this killing power may be tremendously 
multiplied, since it destroys cysts in a dilution as weak as 
1 : 100,000. 


Garcia: Chlorinated Lime and Amoeba 



In order to find the lowest concentration of CaOCl2 suitable 
for internal administration, the writer repeated the same pro- 
cedure and made a series of dilutions of CaOClg from 1 : 1,000,000 
to 5 : 1,000,000, and to every 5 cubic centimeters of the test solu- 
tion a 5 per cent solution of potassium iodide (KI) saturated 
with iodine was added to form a 0.5 per cent solution in the 
various concentrations of CaOCl2. The fsecal material used in 
the previous experiment was utilized. To his surprise he found 
the same structural changes with slight deviations from those 
in Table 1 at different exposures. He found that the lowest 
concentration of CaOCl2 solution with iodine that will kill J57. 
histolytica cysts after one hour fifteen minutes is 4 : 1,000,000. 
The available chlorine (CI2) is about 35 per cent; so that, express- 
ing this in parts per million of CI2, it is 1.4 parts per million 
with 0.5 per cent part of iodine (Table 3). 

Table 3. — Percentages of different types of Entammha histolytica cystg 
observed in different dilutions of chlorinated iodized solutions* 


solution with 

iodine in 0.5 

per cent parts. 

Types of Entamoeba hxitolyiica cysts. 

Control. Original 

faeces with 96 per 

cent «. 4 nuclei, 4 

per cent =» 2 nuclei. 












1 : 1,000.000 

2 : 1.000,000 

8 : 1,000,000 
4 : 1,000,000 
6 : 1,000,000 





I I 
I I 






























No change. 







The mechanism of action of these two chemicals seems to de- 
pend on the production of two nascent halogens of chlorine 
(CI2) and iodine (I2), which are believed to have strong pene- 

308 The Philippine Journal of Science 1935 

trating power through the cyst walls. The nascent CI2 of 
CaOCl2 liberates I2 from KI, and this nascent iodine (I2) plus 
free I2 and CI2 are perhaps enough to kill the cysts. We must 
not forget that the chloramines formed seem to act also on the 
cysts, and these help the two halogens in their killing power. 
The writer can think of no other explanation to account for 
this killing power. 

The above results naturally suggest the possibility of using 
this solution in the treatment of the "cyst-carrier state," which 
Craig (1932) defined as a stage of amoebiasis where symptoms 
may be present or absent and, if present, mild and atypical, 
the condition being characterized by normal or frequent bowel 
movements with abundant cysts in the stool. It is true that a 
large percentage of the cysts in the colon are found in the lumen 
mixed with fseces, and a small percentage in the crypts of the 
mucosal folds. In the selection of the chemical to kill these 
cysts, the one that will act directly on the cysts and is, if 
possible, nonirritating, easily absorbed, excreted by the system, 
and effective in small doses should be chosen. Considering the 
concentration of calcium hypochlorite (CaOCl2) and iodine in 
this mixture, 1.4 parts per million of available chlorine (CI2) 
is suitable for human use and 1 per cent of iodine is the average 
dose of the United States Pharmacopoeia. In this solution the 
writer has only 0.5 per cent iodine, which is 50 per cent less 
than that permitted by the pharmacopoeia and his experience 
as well as that of others shows that 0.5 per cent iodine in wa- 
ter enemata produces no toxic effects, except a mild irritation. 
It was shown also by Sollmann (1917) that iodine is easily ab- 
sorbed in the colon and easily excreted by the system, especially 
by the salivary glands and kidneys. Furthermore, many drugs 
in the market contain iodine. Some of them, as summarize'' by 
Muhlenns (1921), are iodoxyquinoline sulphate, iodochloroxy- 
quinoline, sodium iodoxyquinoline sulphonate (chinioform), iodo- 
chloroxyquineline (vioform), and the popular Yatren 105. 
While all these drugs contain iodine as the basic principle, and 
while various authors claim them to be efficacious, the writer 
believes that chlorinated iodized solution can be used with ad- 
vantage, because of its low cost and easy preparation. Although 
the efficacy of this solution has not been tried in vivo, it certainly 
holds promise in the treatment of cyst carriers. 

*•' ^ Garcia: Chlorinated Lime and Amceba 309 


It may be inferred from the results of the foregoing experi- 
ments — 

1. That the structural changes based on necrobiosis of the 
cell constitute an accurate and reliable criterion of the viability 
of Entamoeba histolytica cysts. 

2. That calcium hypochlorite (CaOCl2) solution is destructive 
to E. histolytica cysts even in as low a concentration as 1 : 100,000 
or 3.5 parts per million of chlorine. 

3. That the presence of nitrogenous organic matter enhances 
greatly the destructive power of calcium hypochlorite (CaOCl2) 
to E, histolytica cysts. 

4. That the concentration of 3.5 parts per million of chlorine 
(CI2), which is effective in killing E. histolytica cysts, is not 
suited for human use and cannot be recommended to free drink- 
ing water from entamoeba cysts. 

5. That solutions of 4 : 1,000,000 calcium hypochlorite (CaO- 
CI2) and 0.5 per cent iodine are also lethal to E. histolytica cysts, 
and may be tried in the treatment of cyst carriers. 


The writer is indebted to Dr. Candido M. Africa, head of the 
Department of Parasitology, School of Hygiene and Public 
Health, University of the Philippines, for his suggestions and 
criticisms during the progress of this work, and for his patience 
in reading the manuscript. 


BOECK, W. C, and Drbohlav. The cultivation of Bntameba histolytica. 
Am. Journ. Hyg. 5 No. 4 (1925) 371. 

BusWMiL, A. M. The Chemistry of Water and Sewage Treatment: A Mo- 
nograph Series. The Chemical Catalogue Co. New York (1928). 

Craig, C. F. The pathology of amebiasis in carriers. Am. Journ. Trop. 
Med. 12 (1932) 285-299. 

DoBBLL, C. The Amoeba Living in Man: A Zoological Monograph. Lon- 
don (1919). 

KUENBN, W. A., and M. H. Sweilmjngrebel. Die Entamoeben des Men- 
schen und ihre praktische Bedeutung. Centralb. fiir Bac. 1. Abt. Orig, 
71 (1913) 78. 

Ll^KB, C. D. Chemotherapy of amebiasis. Journ. Am. Med. Assoc. 98 
No. 3 (1932) 195-198. 

29049 6 

310 The Philippine Journal of Science 

Pinfold, W, J,, H. M. Woodcock, and A. H. Drew. The excystation of 
E. histolytica as an indication of the vitality of the cysts. Brit. Med. 
Jonra. 1 (1916) 714. 

Sbllards, a. W., and M, Thbiler, Investigation concerning amoebic dy- 
sentery. Am. Joum. Trop. Med. 4 No. 3 (1924) 309. 

SOLliMAN, T. H. Textbook of Pharmacology, W. B. Saunders & Co., 
Philadelphia (1917). 

Wbnyon, C. M., and F. W. O'Connor. Human Intestinal Protozoa of the 
Near East. London (1917). 

Wight, T., and V. Wight. On the viability of cysts of E. histolytica under 
variable conditions. Am. Journ. Trop. Med. 12 (1932) 381-886. 

YOBKiSy W., and A. R. D. Adams. Observations on E. histolytica. Ann. 
Trop. Med. and Parasit. 20 No. 3 (1926) 279. 

Yorice:, W., and A. R. D. Adams. Longivity of the cysts in vitro, and their 
resistance to heat and to various drugs and chemicals. Ann. Trop. 
Med. and Parasit. 20 No. 4 (1926) 817. 


[AH figures are camera-lucida drawings of Entamoeba histolytica cysts in the fresh state 
and after treatment with various dilutions of chlorinated lime solutions. X 1,800.] 

Plate 1 

Pioa 1 and 2. Tetranucleate and binucleate cysts in the fresh state in 


3 to 6. Various stages of degenerated cysts under type I. 

7 to 10. Various stages of degenerated cysts under type II. 

11 to 14. Various stages of degenerated cysts under type III. 

15 to 18. Various degenerated and dead cysts under type IV. 


Garcia: Chlorinated Lime and Amoeba.] 

[Philip. Journ. Sci., 56, No. 3 



By Elmer D. Merrill "^ 

Director-in-Chieff New York Botanical Garden 


Curator, Philippine National Herbarium, Bureau of Science, Manila 


While the junior author was studying some of the specimens 
collected in 1930 in the Batan and the Babuyan Islands by 
Messrs. Gregorio Edafio and Maximo Ramos (deceased), botani- 
cal collectors of the Bureau of Science, his attention was called 
by Mrs. Mary S. Clemens to the possibility of one of the speci- 
mens (flowering) being Styrax. More critical study of other 
specimens on hand proved this to be true. 

Several botanical expeditions have been made to the Batanes 
and the Babuyanes, which form the northern portion of the 
Philippine Archipelago, and a report on a collection of plants 
from these islands was made by E. D. Merrill.^ The Batanes 
and the Babuyanes are separated from Formosa by Bashi Chan- 
nel. All of these expeditions brought some plants which were 
new to the Philippines, but known from Formosa. More inten- 
sive explorations of these islands will yield further data on the 
floristic relationships of this group of islands with Formosa. 

The family is new to the Philippines except for F. Villar's 
erroneous record of Styrax benzoin Dry. as a Philippine plant, 
Novis. App. (1880) 127. 


Genus STYRAX Linnaeus 

STYRAX PHILIPPINBNSIS Merr. and Qnis. «p. hot. 

Frutex 2 ad 3 m altus, ramosus, ramis teretibus, glabris, 
ramulis gracilibus, pubescentibus. Foliis brevissime petiolatis, 
ovatis vel ovato-lanceolatis, acuminatis, 4.5 ad 9.5 cm longis, 
2 ad 6 cm latis, supra glabris, subtus parce stellato-pubescen- 

» Philip. Journ. Sci § C 3 (1908) 885-442, map. 


314 The Philippine Journal of Science wsi 

tibus, margine obscure serrulate vel subintegro, nervis utrinque 
4 vel 5. Racemis 2- ad 3-floris, breviter pedunculatis, 10 ad 
13 mm longis, dense pubescentibus. Calycibus campanulatis, 
4 vel 5 mm longis, extus stellato-pubescentibus, intus glabris, 5- 
lobatis, lobis late triangularibus, 0.5 ad 1 mm longis, apice 
truncatis vel rotundatis, minute apiculatis. Corolla campanu- 
lata; petalis utrinque dense adpresse stellato-hirsutis, oblongo- 
elliptieis, obtusis, 16 ad 17 mm longis, 7.5 ad 8.5 mm latis. 
Staminibus 10, aequalibus, 2-seriatis, 9 ad 10 mm longis, fila- 
mentis 5 ad 6 mm longis, sursum glabris, deorsum tomentosis. 
Ovario globoso, tomentoso, stylo 15 ad 20 mm longo, apice trun- 
cato. Fructibus ellipsoideis, 12 ad 13.5 mm longis, 8 ad 10 mm 
diametro, ad basim irregukriter fissis, 1-locellatis, apice ros- 
tratis. Seminibus oblongis, 10 ad 11 mm longis, 5 ad 6 mm 
diametro, 3-sulcatis. 

A branching shrub, 2 to 3 m high; stems terete, glabrous, 
reddish brown or nigrescent, the branchlets slender, pubescent. 
Leaves very shortly petioled, subchartaceous, ovate to ovate- 
lanceolate, acuminate, 4.5 to 9.5 cm long, 2 to 6 cm wide, upper 
surface glabrous, reddish brown or nigrescent, the lower surface 
paler, stellate-pubescent, margins obscurely serrulate or sub- 
entire, the lateral nerves very slender, 4 or 5 on each side of 
the midrib ; petioles 4 to 6 mm long. Racemes 2- or 3-flowered 
at the ends of the branchlets. Flowers white, pedicels 10 to 13 
mm long, densely pubescent. Calyx campanulate, 4 to 5 mm 
long, sparingly stellate-pubescent without, glabrous within, 5- 
lobed, the lobes broadly triangular, 0.5 to 1 mm long, apex trun- 
cate or rounded, with a minute apiculum. Corolla campanulate; 
the petals densely adpressed stellate-hirsute on both surfaces, 
oblong-elliptic, obtuse, 16 to 17 mm long, 7.5 to 8.5 mm wide. 
Stamens 10, equal, 2-seriate, 9 to 10 mm long; filaments some- 
what flattened, 5 to 6 mm long, the upper portion glabrous, the 
basal part tomentose ; anthers linear. Ovary globose, tomentose, 
1-celled; style 15 to 18 mm long; stigma truncate. Capsule 
ellipsoid, apex rostrate, 12 to 13.5 mm long, 8 to 10 mm in dia- 
meter. The pericarp ligneous, coriaceous, straw-colored, when 
dry having a tendency to split irregularly into three valves. 
Seeds oblong, flat on one side, brownish, 10 to 11 mm long, 5 
to 6 mm in diameter, 3-sulcate. 

Philippines, Camiguin Island, Mount Malabsing, Bvr. Sci. 
79248 Edam (type), March 9, 1930 (flowering specimen). 

56,3 Merrill and Quisumbing: Styrax Philippinensis 316 

Batan Island, Mount Matarem, Bur. Set, 80J^i Ramos, June 
20, 1930 (fruiting specimen) . 

A species manifestly allied to Styrax japonicus Sieb. and 
Zucc. and to S. kotoensis Hayata, of Formosa, and should, we 
believe, be considered to be a northern type in the Philippine 
flora. It differs from S. kotoensis Hayata, the flowers of which 
have not been described, in its smaller leaves. 


Plate 1 

Styrax philippinensis Merr. and Quis. sp. nov.: 1, Habit of flowering branch, 
X 0.5; 2, mature flower, x 1.5; S, mature flower with two petals and 
part of calyx removed showing attachment of five stamens and pistil, 
X 1.5; 4, petal, flattened, X 2.5; 5, stamen, X 2.5; 6, fruit, x 1.5; 
7, seed, X 1.5. 


Merrill and Quisumbing : Styrax Philippinensis.] [Philip. Journ. Sci., 56, No. 3 



By Agustin F. Umali 
Of the Fish and Game Administration, Bureau of Science, Manila 


This paper is a systematic description of three rare Philippine 
fishes based on seven specimens collected from the Manila mar- 
kets and Lingayen Gulf, Pangasinan Province, mixed with the 
catches of the beam trawls. They form a part of the ichthyo- 
logical collection of the Fish and Game Administration, Bureau 
of Science. 



Body elongate, cuboid anteriorly and somewhat compressed 
posteriorly, widest and usually deepest at occiput. Either naked 
or covered with very small, cycloid adherent scales arranged in 
very oblique series running downward and backward; those on 
belly inconspicuous or obsolete. Head large, flattened above, 
below, and on sides, almost completely cuirassed with bony plates. 
Eyes small, on upper anterior portion of head. Mouth vertical 
with strong and prominent mandible; villiform teeth on jaws, 
vomer, and palatines; with or without pointed and conical ca- 
nines ; no molars. Maxillary broad, without supplemental bones, 
not slipping under preorbital. Lateral line continuous, running 
high. One or two dorsal fins, spinous portion usually much 
less developed and shorter than soft; anal developed similar to 
soft dorsal; caudal not forked; pectorals large with broad 
oblique bases, lower rays rapidly shortened; ventrals jugular, 
close together, with one spine and five rays. Gill openings wide, 
continued forward; gill membranes nearly separate, free from 
isthmus. Branchiostegals 6; air bladder absent. 

Carnivorous fishes, living at shore bottoms of tropical seas, 
northwards to the Mediterranean and Japan, southwards to 
South Australia and New Zealand. 


320 The Philippine Journal of Science nz% 

Genns URANOSCOPUS Linnttus 

Urano8copu8 Linnjeus, Syst. Nat. ed. 10 (1758) 250. 
Body more or less cylindrical, covered with small and rudi- 
mentary scales. Head large, broad, almost wholly covered with 
bony plates, occipital plates extending to orbits. Opercles and 
shoulder bones usually armed. Grenerally a rounded orifice 
above opercle besides posterior gill opening. Eyes on upper 
surface of head. Cleft of mouth vertical, generally with a 
filament below and before tongue. Villiform or cardiform teeth 
on jaws, vomer, and palatines; no canines. Two separate dor- 
sals, the first with from three to five spines; ventrals jugular. 
Lateral line continuous, running high. Air vessel absent. Pylo- 
ric appendages in moderate numbers. 

URANOSCOPUS JAPONICUS Honttuyn. Plate 1, flff. 1. 

Uranoscopus japonicus Houttuyn, HoU. Maats. Wet. Harlem (1782) 
311; Jordan and Snyder, Proc. U. S. Nat. Mus. 23 (1901) 745. 

Uranoscopus asper Temminck and Schlegel, Fauna Japonica, Poiss. 
(1842) 26, pi. 9, %. 1; GtJNTHER, Cat. Fishes 2 (1860) 226; Jor- 
dan and Snyder, Proc. U. S. Nat. Mus. 23 (1901) 369. 

Dorsal III-IV, 14-15; anal 13-14; pectoral 17-18; scales in 
transverse series about 60. 

Body rather elongate, broad anteriorly and compressed toward 
caudal peduncle; depth 3.6 in length. Head flat above and on 
sides, with moderately rugose bony plates, dorsal plate extend- 
ing forward between eyes where it is divided by a deep U- 
shaped, naked depression; preorbital plate with a flat spine; 
preopercle with four short spines on lower edge and another 
stronger one at anterior border of subopercle; humeral spine 
sharp and long, about 3.2 in length of head,^ a short flat spine 
above this. Head 3 in body length; greatest width 1.4 in length 
of head. Snout very short, 7 in head; mouth vertical; max- 
illary 2.1 in head; teeth conical and. movable in two or three 
rows in upper and lower jaws, in villiform bands on vomer, 
palatines, and pharyngeals. Eye small, on upper side of head, 
5.8 in head, its horizontal diameter equal to interorbital. Gill 
rakers on first arch obsolete, represented by a few small 
bunches of setse. 

Scales small, square, deeply embedded ; arranged in about sixty 
oblique rows ; occiput, breast, and belly, as well as a narrow strip 
along base of dorsal, and a narrow area along base of anal fin 

*The length of head includes the lower jaw in the description of the 
species of this family. 

56,3 Umali: Philippine Fishes 321 

naked; lateral line running high, extending along base of dorsal 
fin abruptly descending to middle of base of caudal at region 
of caudal peduncle. Dorsals separate, the first with three or 
four short flexible spines, the highest about 3.5 in head, and 
the last hardly visible in most specimens as in Plate 1, fig. 1; 
anterior rays highest, 1.9 in head. Anal inserted below first 
ray of second dorsal, its fin membrane and especially branched 
portion of rays fleshy; caudal convex posteriorly, middle rays 
of pectoral longest; ventrals about 2 in head, three pairs of 
strong spines directed forward projecting from their bases. 

Color of fresh dead specimen, uniform dark brown above, 
fading to grayish white towards sides, white on belly; spinous 
dorsal black, basal fin membrane white; soft dorsal and caudal 
yellowish gray, the former with dusky spots along rays; anal 
dusky posteriorly, other portions yellowish with white edge; 
pectorals yellowish with narrow white margin; ventrals white, 
the edge dusky. 

Color in alcohol brownish, fading to whitish toward belly. 
Spinous dorsal black, with narrow white band at base, the spots 
along fin rays totally faded; all other fins yellowish, anal and 
ventrals paler. 

Description based on five specimens. No. 31250, total length 
150 millimeters, obtained from Quinta Market, Manila, October 
25, 1930; Nos. 41063, 41064, 41065, and 51565, 117, 108, 128, 
and 154 millimeters in total length, respectively; all collected 
from Lingayen Gulf, Pangasinan, October 3, 1934. 

URANOSCOPUS BICINCTUS Temminck and Schlearel. Plate 1, fig, 2. 

Uranoscopus bicinctus Temminck and Schlegel, Fauna Japonica, 
Poiss. (1842) 26, pi. 10, B; Bleeker, Act. Soc. Nederl., II, Am- 
boina (1857) 411; Gunther, Cat. Fishes 2 (1860) 228; Nystrom, 
Svensk. Vet. Akad. (1887) 28; Ishikawa, Prel. Cat. (1897) 46. 

Dorsal IV, 13 ; anal 13 ; pectoral 18 ; scales in transverse series 
about 75. 

Body robust anteriorly and compressed posteriorly, broader 
than deep on head; depth 4 in length. Head flat above, covered 
with coarsely granular bony plates throughout except a narrow 
U-shaped naked portion at interorbital ; preorbital with a flat 
blunt spine; preopercle with four short spines directed down- 
ward on lower edge, another stronger spine at anterior border 
of suboperculum ; humeral spine short and stout, about 5 in head, 
two short spines above and anterior to this, at superior angle 
of operculum. Head about 2.7 in body length; greatest width 

322 The Philippine Journal of Science 1935 

about 1.2 in length of head. Snout very short, 6.4 in head; 
mouth vertical; maxillary 2.6 in head; teeth conical and mov- 
able, in narrow bands in upper and lower jaws, in villiform 
bands on vomer, palatines, and pharyngeals. Eye prominent 
on upper side of head, 7.7 in head, its horizontal diameter con- 
tained 1.4 in interorbital. Gill rakers on first arch obsolete, 
represented by a few small bunches of setse. 

Scales small, square, deeply embedded; arranged in about 
seventy-five oblique rows ; naked at occiput, breast, and belly, as 
well as a narrow strip along base of dorsal and anal. Lateral 
line extends along base of dorsal fin, curving downward to mid- 
dle of caudal fin at region of caudal peduncle. Dorsals separate, 
first dorsal of four short, flexible spines, the longest 4.5 in 
head; second dorsal of thirteen rays, third ray longest, 2.2 in 
head. Anal inserted below origin of second dorsal; fin mem- 
brane, especially at branched portion of rays, fleshy; caudal 
somewhat convex; middle rays of pectoral longest; ventrals 
2.1 in head, three strong spines directed forward projecting from 
base of fin. 

Color in alcohol, blackish brown above, fading to grayish white 
towards lower sides and belly; head and sides of body speckled 
with minute dark dots; a broad, dark blotch (somewhat in- 
distinct in my specimen) passes around body through base of 
first dorsal, and an obscure black spot below posterior base of 
second dorsal. First dorsal black, with a white spot at anterior 
and posterior of base. Second dorsal dusky, especially towards 
margins, dotted with black. All other fins dusky yellowish with 
small black spots; anal, pectorals, and ventrals with white edges. 

Here described from a single specimen. No. 31249, total length 
262 millimeters, obtained from Divisoria Market, Manila, Jan- 
uary 10, 1931. 

Found on the coasts of Japan and southward. 


Body elongate, compressed, more or less eel-shaped; either 
naked or covered with very small cycloid scales embedded in 
skin and arranged in oblique series at right angles to each other. 
Head large, naked or partly scaled ; lower jaw included. Mouth 
moderately protractile, both jaws and generally vomer and 
palatines also, with villiform or cardiform teeth. Barbels pres- 
ent or absent. Eyes moderate in size, rudimentary or entirely 

»«'« Umali: Philippine Fishes 323 

wanting. Lateral line more or less distinct. Vertical fins of 
soft rays only; confluent with each other. Pectoral fins some- 
times wanting. Ventral fins each reduced to a pair of filaments 
or a bifid ray inserted just behind chin between rami of lower 
jaw. Gill openings very wide; gill membranes separate, attached 
to isthmus behind ventrals. Pseudobranchiae small; gills 4, a 
slit behind fourth. Air bladder and pyloric caeca present. 

Small, marine carnivorous fishes found in most warm seas, 
some descending to considerable depths in the Atlantic, Pacific, 
and Indian Oceans. 


Leptophidium Gill, Proc. Acad. Nat. Sci. Phila. (1863) 210. 
Lepophidium Gill, Am. Nat. 29 (1895) 167. 

Body elongate, more or less compressed, decreasing gradually 
in height towards posterior where it ends abruptly in a point; 
scales regularly imbricate in quincunx. Head with imbricated 
scales extending to forehead; snout high, obtusely rounded, and 
projecting forward, with or without a short, nearly concealed 
spine directed forward and somewhat downward; mouth mod- 
erate, oblique; teeth in jaws villiform, embedded in a mucous 
membrane, separated by an interval from longer and pointed 
ones in outer row; vomer and palatines with teeth. 

LEPOPHIDIUM MARMORATUM Goode and Bean. Plate 1. fiff. 3. 

Leptophidium marmoratum Goodb and Bean, Proc. U. S. Nat. Mus. 7 
(1885) 423; Goodb and Bean, Oceanic Ichthyology (1895) 348, fig. 

Dorsal about 90 ; anal about 60 ; series of scales along- lateral 
line about 86. 

Body more or less elongate and eel-like, moderately com- 
pressed anteriorly and ribbonlike posteriorly, and gradually 
tapering to a point at tail ; depth 5.6 in length. Head scaled on 
cheeks and forehead, 4.5 in body length; snout blunt, without 
spines, projecting forward over mandible, 4.1 in head, equal to 
interorbital. Mouth moderate, oblique with villiform teeth on 
jaws, vomer, and palatines, outer series larger and movable; 
maxillary extending slightly beyond vertical through posterior 
margin of orbit, its length equal to postorbital portion of head. 
Eyes lateral and moderate, 3.2 in head, somewhat exceeding 
snout. Opercle ending in a small spine somewhat hidden in 
opercular flap. 

Scales small, cycloid, closely imbricate, about eighty-six along 
lateral line; lateral line apparently complete. Dorsal low, of 

324 The Philippine Journal of Science 

about ninety soft rays, confluent with caudal and anal; anal 
of about sixty soft rays. Pectoral fins short, the length 2 
in head. Branchiostegals seven; pseudobranchise present; gill- 
rakers short, eight below angle of first arch^ the longest about 
0.3 diameter of eye. 

Color in alcohol yellowish, lighter towards belly and under- 
surface of head. A broad brown blotch on forehead continued 
downward as an oblique band along posterior border of eye 
to lower edge of opercle; two brown bands on nape before dorsal 
origin running obliquely downward to belly, the posterior one 
sending a branch as a longitudinal stripe along whole length 
of lateral line; another brown stripe along base of dorsal fin. 
Anterior half of dorsal fin white with irregular blackish to 
black blotches ; posterior half of dorsal, caudal and anal blackish 
brown, base and outer edges white. Pectorals and ventrals 
whitish, base of former brownish or dusky. 

Description based on a single specimen. No. 31248, total length 
140 millimeters, obtained from Divisoria Market, Manila, 
November 26, 1930. 


[Drawings by Angel D. Lagman.l 

Plate 1 

Fig. 1. Uranoscopus japonicus Houttuyn. 

2. Uranoscopus hicinctus Temminck and Schlegel. 

3. Lepophidium marmoratum Goode and Bean. 

290499 ^7 325 


By M. W. DB Laubenfels 

Of Altadena, California 


Prof. Hilario A. Roxas, formerly of the University of the 
Philippines, has transmitted to me for identification a small but 
interesting collection of sponges. These were all collected near 
the marine biological station of the University of the Philip- 
pines, v^hich is situated at Puerto Galera Bay, Mindoro Island, 
Philippine Islands. The sponges are exclusively from very 
shallow water, in no case more than 12 meters deep. Two of 
the nine species that could be identified proved to be new, which 
is very remarkable. It is earnestly hoped that more work will 
be done with the Porifera at this marine station. Several of 
the forms in question should prove very interesting for phys- 
iological experimentation. 

None of them belongs to the class Calcispongiae, and since 
this collection is exclusively from shallow waters, none is of the 
class Hyalospongise. The nine species represent nine genera of 
the class Demospongise. 

Probably the most extensive treatise on the sponges of the 
Philippine regions is that by H. V. Wilson.^ 


The Puerto Galera sponge here described is a massive to 
irregularly lamellate specimen, gray, tinged with lavender, as 
preserved in alcohol. Its consistency is spongy. There is a 
smooth, presumably horny, dermis, as is true of keratose 
sponges in general, and it is stretched over conules 2 mm high 
and 1 to 8 mm apart. Underneath this dermal membrane there 
probably have been subdermal cavities, but in general these 
are collapsed at the time of examination. The pores and oscules 
are not evident. In the endosome there is an obvious reticula- 
tion of the fibers. The flagellate chambers are eurypyllous, 40 \k 

*BuU. U. S. Nat. Mus. 100 (1925). 


328 The Philippine Journal of Science i985 

by 65 [I., oval. Of the above-mentioned fibers, the mesh, ap- 
proximately square, is about 0.45 to 1 mm in diameter. The 
fibers are about 360 in diameter and are crowded with foreign 
material so that the spongin is scarcely evident. The fascicular 
nature of the specimen studied by Wilson is barely in evidence, 
but there can be little doubt that it is the same variety as the 
one found by that author near this locality. 

This specimen is in general quite characteristic of pallescens, 
especially in the lavender color and the wide distribution of the 
conules. The other conspicuous member of the genus Dysidea 
is D. fragilis, which is a pale gray sponge lacking the lavender 
and having many small conules very close together, which are 
usually less than 1 mm high and less than 1 mm apart, apex 
to apex. 

Remarks, — This species was described as Spongelia pallescens 
by Schmidt (1862, p. 29) in his monograph on the sponges of 
the Adriatic, and referred to Spongelia by Schulze (1879, p. 138) . 
There may be some doubt that the Philippine sponge is con- 
specific with Schmidt's from, the Mediterranean. Wilson (1925, 
p. 476) reported from the Philippine Islands a sponge which 
is clearly conspecific with that in the collection at present under 
discussion, and identified it as Dysidea fragilis var. fasciculata 
Wilson. Wilson explained that he employed the specific name 
in question rather than pallescens because Lendenfeld (1889, p. 
642) had reduced pallescens in synonymy to Spongia fragilis 
of Montagu (1818, p. 114). But this is not a good reason. 
Kallmann ^ demonstrated conclusively that the latter author was 
utterly unreliable in his conclusions and descriptions. 

ADOCIA BAERI (Wilson). Plate 1, flff. 1. 

The Puerto Galera specimen is represented by some frag- 
ments, perhaps all from the same colony, which colony was pre- 
sumably ramose. The fragments are cylindrical, usually hollow, 
and in one case dichotomously branched. The color as pre- 
served in alcohol is dark, almost black. The consistency is very 
soft. The surface, as is very characteristic of haeri, is minutely 
tuberculate. A dermal region of specialized pigment cells ex- 
tends some 400 y^. into the sponge. The endosome is a typical 
"renierid'* or isodictyal reticulation, with conspicuous canals 
about 1 mm in diameter, running longitudinally in the cylinders, 
parallel to the surfaces of the exterior and of the cloacal hollow. 

* A Revision of the Monaxonid Species Described as New in Lendenf eld's 
Catalogue of the Sponges in the Australian Museum (1914). 

*^'3 De Laubenfels: Sponges from Puerto Galera 329 

There are what appear to be pores, remarkably large, even 1 mm 
in diameter in many cases. The oscule or cloacal aperture is 
terminal and 3 mm in diameter. The megascleres are rather 
sharp pointed oxeas, fairly uniform in size, 5 ii by 130 \k. 
Among them are considerable numbers of much smaller oxeas, 
only 0,5 jx by 90 \l. These are perhaps undeveloped megascleres 
of the principal type or, on the other hand, perhaps to be 
regarded as microscleres. Spicules, while not yet fully formed, 
are well known to resemble the fully formed ones in shape, 
except for much greater slenderness, and the presence of such 
spicules must therefore have very doubtful taxonomic value. It 
is always possible that they may represent a special category, 
but it can seldom be ascertained to any degree of satisfaction 
whether or not this is the case. 

Remarks. — ^Wilson (1925, p. 398) reported from the Philip- 
pine Islands a specimen obviously conspecific with the present 
variety, which he identified as Reniera implexa Schmidt var. 
baeri Wilson (1868, p. 27). Burton^ demonstrates the suit- 
ability of the genus Adocia rather than Reniera for cases such 
as the one under consideration; furthermore, the differences 
between the Philippine and the Mediterranean forms are suf- 
ficiently great for full specific rank. Within the genus under dis- 
cussion there are so few criteria upon which to base specific 
differences that items which would appear trivial in other groups 
must here receive major attention. 

GELLIODES FIBROSA (Wilson). Plate 1, fig. 2, 

The Puerto Galera species here described is a small cavernous 
mass represented in places by only a series of trabeculse each 
about 2 by 4 mm. As preserved in alcohol it is blue-gray on 
the exterior over a paler gray not at all blue on the interior. 
The consistency is quite spongy. The ectosomal specialization 
is a membranous sheet, transparent, and contains tangentially 
placed spicules. The endosome is very cavernous, with cham- 
bers 6 mm in diameter, elongate, meandering through the inte- 
rior. The architecture is pronouncedly fibrous as is character- 
istic of the genus GelUodes but not of the so-called Gelliiis. 
These fibers contain many spicule rows, and range from 60 [i. to 
160 |x in diameter. The mesh that they form is commonly 700 \k 
to 2,000 \L in diameter. There are auxiliary connectives which 
are fibers of only about half (or less) the diameter of the 

* Sponges of the Great Australian Reef (1934). 

330 The Philippine Journal of Science 1935 

principal ones, and these inclose meshes only 40 [jt. to 60 [a in 
diameter. The spicules are typically oxeas, but perhaps 3 to 
5 per cent of them are styles. Typical sizes are 6 |x by 190 [x, 
5 [A by 175 (A, 5 (A by 150 [x, 4 [jl by 150 [x. Among these are 
sigmas of great variation in size and thickness; the length of 
chord varies from 11 [a to 22 [a, and the thickness from much 
less than to nearly 2 \l. 

The cavernous structure is perhaps the most characteristic 
feature of this species as contrasted with others of the genus 
GelUodes. The spicules are rather smaller than those of most, 
and the lack of conules also is noteworthy. 

Remarks. — This sponge was described from the Philippines 
by Wilson (1925, p. 398) as being conspecific with the sponge 
first described as Halichondria varia by Bowerbank (1875, p. 
292), and later referred to Gellius by Ridley and by other 
authors. Wilson established for the Philippine sponge a new 
variety, fibrosa. The present specimen sheds still further light 
on the nature of the species over and above the information 
available to Wilson, and it now becomes evident that instead of 
being a Gellius (Adocia), this is a GelUodes. Therefore, it 
becomes necessary to treat the name fibrosa as of full specific 
rank instead of as a variety. It is not implied that varia of 
Bowerbank is also a GelUodes, but rather that it is not even con- 
generic with the Philippine sponge. 

LISSODENDORYX ROXASI sp. nor. Plate 1. fi^. 5. 

This Puerto Galera sponge is about 2 cm by 6 cm, an irregular 
reticulation of trabeculse, very cavernous in nature. The color 
is drab, and the consistency is very elastic. Because of the 
cavernous, perhaps partially macerated condition, the pores and 
oscules cannot be described. A large amount of the dermis may 
have been lost. The reticulation made by the structures referred 
to above consists of meshes only some 2 mm apart, center to 
center, and nearly rectangular in outline. The trabeculse them- 
selves are nearly 1 mm in diameter. The ectosomal spicules are 
uncommon, many having perhaps been lost since the sponge 
was preserved. They are tylotes, 3 ix by 220 [x. The abundant 
principal spicules are crowded rather irregularly in the endo- 
some. They are entirely spined styles, 8 \h by 155 ^. Many 
of them are so placed as to echinate the trabeculse. There are 
two size ranges of arcuate isochelas. The abundant ones are 
about 36 (i. long, but a rarer sort is only 16 \l long. Sigmas 70 |i 
in chord length are present, but not at all common. 

W.3 De Laubenfels: Sponges from Puerto Galera 331 

Remarks.—Wilson (1925, p. 432) described a species of Lis^ 
sodendoryx, from the Philippine Islands, naming it tawiensis. 
There is little reason to believe that this species is conspecific 
with the form under consideration. In fact, it is not even 
congeneric, especially since its chelas are not arcuate, but dis- 
tinctly palmate. It is, therefore, here proposed to regard it as 
the genotype of a new genus to be known as Myxichela. The 
diagnosis would be, "resembling Myxilla except for the sub- 
stitution of palmate chelas instead of any other kind." The 
nearest of the numerous species of Lissodendoryx to roxasi 
seems to be that described by Topsent (1897, p. 457) as baculata, 
from the East Indies. Of the latter, the dermal spicules are 
strongyles instead of tylotes, but of about the same size as those 
in roxasi. The principal megascleres are somewhat larger, and 
smooth. It has only the smaller size range of isochelas, and 
its more numerous sigmas are smaller. 

The genus Lissodendoryx is very close to the genus Myxilla 
from which it was originally separated by having smooth prin- 
cipal spicules, whereas Myxilla has spiny megascleres. This 
fact by itself alone proved to be an untenable basis of differen- 
tiation, and it was noted that the genotype has this further 
significant difference, that its chelas are arcuate instead of 
anchorate, as in Myxilla, and this distinction has been recently 
accepted as the firm basis for differentiation. However, there 
are very few species of Lissodendoryx that possess spines upon 
the principal megascleres ; the sponge now being described is one 
of the few presenting this characteristic. 

The species described here is named in honor of Prof. Hilario 
A. Roxas, of the Bureau of Science, Manila. 

MONANCHORA DIANCHORA ap. nor, Plate 1, flir. 4. 

This Puerto Galera sponge was a lamellate mass about 1 cm 
thick, and 4 by 5 cm in area, curled in such a way as almost 
to make a funnel. Preserved in alcohol, it is light orange-red 
on the exterior and paler orange on the interior. The consis- 
tency is soft and spongy. The surface might perhaps be de- 
scribed as conulose, but there are only about a dozen conules on 
the entire surface, which is otherwise smooth. The conules in 
question are about 4 mm high. The dermal membrane is thin 
and can only with great difficulty be detached from the sub- 
dermal cavities which extend beneath it. In it the dermal spic- 
ules are tangentially placed. The endosome is very confused, 
but vague ascending tracts can be made out here and there, 

332 The Philippine Journal of Science 1935 

consisting of approximately eight spicule rows each, and being 
about 40 |x in diameter. The megascleres are of one type only, 
namely tylostyles, but tend to be of two sizes. Those in the ecto- 
some are approximately 4 [jl by 283 \i. Typical sizes for those 
in the endosome may be given as 7 [a by 276 (jl, 5 [a by 370 |x, 
9 t^ by 270 ji.. The microscleres are most distinctive. There 
are two kinds, appearing superficially like sigmas, but being 
really unguiferate chelas having several teeth at each end. The 
conunonest category is of isochelas of this type, 33 [a in chord 
length, and having at each end four sharp-pointed teeth. A 
second category is of similar spicules, only 20 ^ in chord length 
and apparently having only three minute teeth at each end. 

Remarks. — This sponge may be compared with Monanchora 
clathrata Carter (1883, p. 369), the type of the genus, which it 
resembles in general, except that Carter's species had at least 
five teeth at each end of the peculiar chelas. We do not have 
many data pertaining to the Australian species, there being little 
information available, for example, as to its flesh and structure 
in general. Another species of especial interest here is the one 
described by Dendy (1921, p. 58) as Amphilectus unguiculatus, 
from the Indian Ocean. Its megascleres are considerably longer 
but not thicker than those of dianchora, and its microscleres are 
very unusual in their flat, straplike shape, the shaft not being 
round in cross section as might be expected. 

It is interesting to find another representative of this rare 
genus characteristic of the East Indian and Australian regions. 
It is somewhat paradoxical that Carter's choice of a generic 
name was most inappropriate. Whereas the Philippine species 
has more than one kind of "anchor," it is clearly so closely 
related to clathrata that it scarcely seems appropriate to estab- 
lish for it at the present time a new genus, merely because it 
does not conform to the description implied in the generic name. 

DRAGMAXIA CILIATA (Wilac^n). Plato 1, llff. 5. 

The Puerto Galera sponge here described is a lobate mass 
about 7 by 3.5 by 3.5 cm. The surface is smoothly rounded 
into somewhat less than a dozen nearly spherical lobes. The 
color as preserved in alcohol is brown, darker on the exterior, 
and paler in the interior. The consistency, like that of the 
genus Suberites, is corklike. The surface, however, is pro- 
nouncedly hispid, with a spicule fur about 80 ^ high consisting 
of spicules some 20 jjt. apart, placed erect; otherwise the surface 

56,8 De Laubenfels: Sponges from Puerto Galera 333 

may be described as nearly smooth. There are numerous pores 
28 (X to 40 (x in diameter, and about 80 [a to 90 (x apart, but the 
oscules cannot be made out with certainty. The endosome is 
pronouncedly radial, and very dense in structure, lacking the 
breadlike appearance frequently found in sponges of the Suberi- 
tidae. The flagellate chambers are round, presumably diplodal, 
varying from 28 [a to 40 (jl in diameter. The megascleres are 
styles. Those of the ectosome are only about 4 [x by 320 (jl, a 
few being as much as 5 ^i. in diameter, but many only as thick 
as 2 [x. Those of the endosome are much larger, varying fre- 
quently between 22 p. by 930 pt. and 26 [x by 1040 \l. Among 
them are very small spicules or rhaphides less than 1 ix in dia- 
meter and usually a little more than 200 ja in length. 

Remarks. — ^Wilson (1925, p. 341) described this species from 
the Philippines as Tuberella ciliata. Topsent in 1900 in his 
"Etude monographique des Spongiaires de France" made clear 
that Aaptos and Tuberella were synonymous. The former dates 
from Gray (1867) and the latter only from Keller (1881), so 
it is evident that Tuberella should be dropped in favor of Aaptos, 
not vice versa. It appears that Wilson's specimen had few or 
none of the rhaphides, so Dragmaxia was not suggested. 

The genus Dragmaxia has for a long time been represented 
exclusively by its genotype, the sponge described by Whitelegge 
(1907, p. 513) as Spongosorites variabilis. It is quite a pleasure 
to find another species for this rare and interesting genus. 
Whitelegge established the genotype with the understanding 
that his species was an epipolasid, that is to say, a reduced 
tetraxon sponge. Apparently Kallmann (1916, p. 543) was 
quite right when he transferred it (as of his new genus Drag- 
maxia) to the Axinellidae. A still further interesting concept is 
brought out by comparison with the genus Aaptos; Dragmaxia 
ciliata is very much like Aaptos, except for its possession of 
oxeote microscleres. The latter genus is usually regarded as 
suberitid rather than axinellid. The dense subspherical shape 
of ciliata suggests a suberitid relationship, but its hispid ecto- 
some points more forcibly toward the Axinellidae. All items 
considered, this is not only an interesting but also a perplex- 
ing species. 

Comparing this sponge with the other species of Dragmaxia 
mentioned above, that species {variabilis) is lamellate, and 
its principal spicules are nearly double the thickness without 

334 The Philippine Journal of Science isss 

being longer than those of ciliata; its microscleres are about 
five times as thick without being longer. A fairly close rela- 
tionship is indicated. Another genus that is worthy of con- 
sideration here is Alloscleria Topsent, which, however, has two 
types of diactinal microscleres, one kind acanthoxeas, and the 
other centrotylote smooth oxeas. 

BPIRASTRELIA VAGABUNDA Ridley. Plate 1, flff. 6. 

The present Puerto Galera species is a compact mass 4 cm 
in diameter and 6 cm in height, showing pronounced vertical 
grooves on the lateral surfaces, so that it becomes almost digi- 
tate. In alcohol it is a rich brown, and the consistency resem- 
bles that of cork. The color is due to very strongly pigmented 
bodies contained in a layer of cells about 40 [x deep from the 
surface. In general the surface is to be described as smooth. 
No membrane can easily be removed from it. The structures 
in the interior are as confused as are those of the exterior, but 
more of the spicules have their points directed towards the 
surface of the sponge than in any other direction. There are 
no special dermal spicules. The megascleres are of one type 
only, namely tylostyles, of which the following sizes are represent- 
ative: 13 \L by 570 [x, 14 \l by 540 [x, 16 \k by 594 [i, 14 \i by 634 [x, 
9 [JL by 405 \k. The microscleres are much bent or spiral spiras- 
ters with very small spines; the entire spicule averages little 
more than 1 pi. in thickness, and the maximum length is about 
15 [x. 

Remarks, — ^Very numerous species of Spirastrella have been 
described from various parts of the world and they do not 
differ greatly from each other. Vosmaer in 1911 in connection 
with the Siboga reports, published a monograph on this genus, 
in which he reduced a very large number of species in synon- 
ymy to Spirastrella purpurea. It is here considered that Wil- 
son's remarks in the 1925 reference mentioned above are correct; 
namely, that by no means all of the species mentioned by Vosmaer 
should be so reduced. 

The sponge at present under consideration is obviously the 
same as that described by Wilson (1925, p. 343) with the above 
identification. This statement calls for some discussion, but 
little more need be said than what has already been written 
by Wilson in the reference cited. Vagabunda was established by 
Ridley (1884, p. 468), in his report on the Porifera collected by 
the Sea Lark in the Indian Ocean. 

56,3 De Laubenfels: Sponges from Puerto Galera 335 

MYRIASTRA CLAVOSA (Ridley). Plate h fig. 7. 

This Puerto Galera sponge is represented by a number of 
small spherical masses some 6 to 8 mm in diameter each. The 
color as preserved in alcohol is very pale brown, almost white, 
and the consistency is rather elastic but similar to cork. The 
cortical specialization is very thin for a tetraxon sponge of this 
nature, being only from 80 [x to 95 (a thick. It extends over 
very definite subdermal cavities. The endosomal structure is 
pronouncedly radial. Pores cannot be distinguished with cer- 
tainty, and each of the small sponges seems to have just one 
oscule, which is about 1 mm in diameter. The principal spicules 
are large dichotrisenes with their cladomes in or just below the * 
cortex. They are frequently as long as 4 mm. The rhabds are 
about 50 mm in diameter, and the chord ranges from 570 (x to 
660 [A. There are anatriaenes, protrisenes, and orthotrisenes, as 
well as oxeas. The rhabds of the former, and the total diameter 
of the latter, are in the neighborhood of 18 \l. The chords of 
the trisenes range from about 65 [jl to nearly 100 [x. Microscleres 
seem to be of one sort only, namely tylasters. They vary 
from 9 [x to 15 [x in diameter, and (as usual among the tetraxon 
sponges) the smaller asters have the more numerous rays, up 
to fourteen or fifteen, whereas the larger have less numerous 
rays, only three to five. 

Remarks, — This species was first described as Steletta clavosa 
by Ridley (1884, p. 474), in his report of the Indian Ocean 
sponges collected by the Sea Lark. A very complete and satis- 
factory discussion of it is given by Wilson (1925, p. 387). 


The Puerto Galera sponge here described is a subspherical 
mass about 1 cm in diameter, light brown in color, and with a 
stony hard cortex. Underneath this the endosome is brittle 
but somewhat like cork in other ways. These characters are 
quite typical of the genus Geodia, The cortex is about 800 \l 
thick, and is densely packed with sterrasters as characteristic 
of the group. The pores and oscules cannot be distinguished. 
The endosome is pronouncedly radial in structure. The prin- 
cipal megascleres are plagiotrisena with rhabds about 20 [x by 
1685 IX (a few being much longer), and with chords 265 [x 
across. In this specimen there occur also orthotrisenes of 
which the rhabds appear always to have been broken. The clads 
are about 350 [x to 430 \i long. The sterrasters are approxi- 

336 The Philippine Journal of Science 

mately 108 ^ by 121 p., and, of course, developmental forms there- 
of occur. The euasters appear to have their arms microspined ; 
they reach a diameter of approximately 12 ^ to 16 ^. There 
are furthermore rather numerous spherasters, chiefly centrum, 
having a diameter of barely 4 ^. 

Remarks. — This sponge was described as a new species from 
the Philippines by Wilson (1925, p. 314). 


Plate 1 

[From camera-lucida drawings. Figs. 1 to 4 and 7, e are X 630 ; figs. 5, 6, and 7, a to d 

are X 150.] 

Fig. 1. Adocia baeri (Wilson) ; a, oxeote megasclere; 6, rhaphide. 

2. Gelliodes fibrosa (Wilson); a and b, oxeas; c, sigmas. 

3. Lissodendoryx roxasi sp. nov.; a, tylote; 6, acanthostyle ; c, sigma; 

d, larger (arcuate) chelas; e, smaller chelas. 

4. Monanchora dianchora sp. nov.; a, tylostyle; 6, larger (unguiferate) 

chelas; c, smaller chelas. 

5. Dragmaxia ciliata (Wilson) ; a, endosomal style; 6, ectosomal style; 

c, rhaphide. 

6. Spirastrella vagabunda Ridley; a, tylostyle; 6, spirasters. 

7. Myriastra clavosa (Ridley) ; a, cladome of dichotriaene; b, cladome 

of plagiotriaene ; c, protriasne; d, cladome of anatriaene; e, tylas- 


De Laubenfels: Sponges from Puerto Galera.] 

[Philip. Journ. Scl, 56, No. 3 


f=^<r% ^ 










By Charles P. Alexander 
Of Amherst, Massackuaetta 


The crane flies discussed in this paper are chiefly derived from 
three sources; namely, Celebes, collected by Mr. Charles F. 
Clagg; Formosa, by Mr. J. Linsley Gressitt; and Kashmir, by 
Miss Vivien R. Hutchinson. Scattered additional species were 
secured in Java by Dr. Ian Mackerras; in Formosa by Prof. 
Syuti Issiki; and on Mount Omei, western China, by the Rev. 
Mr. George M. Franck. One further species, from eastern Si- 
beria, was included in the rich collections belonging to the Rus- 
sian Academy of Sciences, Leningrad, sent to me for study by 
Dr. A. von Stackelberg. I wish to express my deepest thanks 
to all of the above-named entomologists for their continued 
kindly cooperation in my study of the Tipulidae of Asia. Ex- 
cept where indicated to the contrary, types of the various novel- 
ties are preserved in my collection. As has been done in other 
parts under this series of reports, I am including a species of 
Trichoceridae, preferring this course to preparing a special paper 
on this insect. 


TRICHOCERA ARISANENSIS sp. nov. Plate 1, &«. 1, 

General coloration of thorax brownish yellow, the prsescutum 
with three darker brown stripes; pleura chiefly dark brown, 
variegated by paler brown; femora brown, their bases obscure 
yellow; wings yellowish gray, sparsely patterned with brown; 
abdominal segments conspicuously bicolorous, the basal portion 
brown, the outer ring light yellow. 

Female. — Length, about 6 millimeters; wing, 6.5. 

Rostrum dark brown; palpi elongate, brownish black. An- 
tennae black, the scape and pedicel paler ; outer flagellar segments 

^Contribution from the entomological laboratory, Massachusetts State 


340 The Philippine Journal of Science 1935 

paler than the basal ones; flagellum much more elongate and 
much slenderer than in pictipennis, the individual segments 
correspondingly lengthened. Head brown. 

Mesonotal prsescutum brownish yellow with three darker 
brown stripes; scutal lobes dark brown, the median area more 
yellowish; scutellum and mediotergite more testaceous brown. 
Pleura chiefly dark brown, variegated by paler brown on the 
pteropleurite and posterior anepisternum ; dorsopleural mem- 
brane yellow. Halteres pale yellow, the knobs dark brown. 
Legs with the coxse and trochanters yellowish testaceous; femora 
obscure yellow basally, passing into brown outwardly; tibise and 
tarsi brown. Wings (Plate 1, fig. 1) with the ground color 
yellowish gray, sparsely patterned with brown, including areas 
at origin of Rs, along cord, on vein R2, outer end of cell 1st 
M2, and the fork of vein Mi+g; paler and less evident areas in 
the outer radial field ; no darkening at end of vein 2d A ; veins 
brown. Venation : R2+8+4 a little shorter than Ri+g ; Sci ending 
about opposite fork of R2+8-H; cell Mi subequal to or a little 
longer than its petiole; m-cu shortly before fork of Ma^^; vein 
2d A somewhat angular, at extreme outer end bent suddenly 
into margin. 

Abdominal segments conspicuously bicolorous, the basal two- 
thirds of the individual segments brown, the outer third light 
yellow. Cerci pale yellow, the tips acute. 

Habitat, — Formosa. 

Holotype, female, Arisan, altitude 7,640 feet, May 24, 1934 

The nearest allied species is Trichocera pictipennis Alexander, 
of Japan and Korea, which differs conspicuously in the colora- 
tion of the thorax and abdomen, in the shorter antennae, and 
in the details of the wing pattern. The genus Trichocera is new 
to the fauna of Formosa. 


LONGURIO (LONGURIO) VARICEPS Bp. nor. Plat* 1, fig, 1. 

General coloration of thorax black; head fulvous-oirange, with 
a conspicuous black longitudinal stripe on either side of vertex; 
halteres and legs black throughout; wings strongly tinged with 
black, the costal portion and seams along vein Cu and anterior 
cord still darker; Rs short, a little less than cell 1st M2; abdomen 
with segments six to nine, inclusive, black; tergites two to five 

»6, 3 Alexander: Tipulidse from Eastern Asia, XXIII 341 

orange, narrowly bordered by blackish; basal sternites con- 
spicuously bicolorous, black, variegated by yellow. 

Male. — ^Length, about 21 millimeters; wing, 16.5; abdomen 
alone, 16.5. 

Female.— Length, about 25 millimeters; wing, 18; abdomen 
alone, 20. 

Rostrum and palpi velvety black. Antennae black through- 
out; outer verticils very long and conspicuous. Head dull 
fulvous-orange, with a conspicuous black longitudinal stripe 
extending from either side of the vertical tubercle along the 
orbital region, behind becoming more diffuse and involving the 
occipital region, leaving the median area of the posterior vertex 
of the ground color. 

Thoracic dorsum and pleura entirely deep velvety black. Hal- 
teres and legs black throughout. Wings (Plate 1, fig. 2) very 
strongly tinged with black, the prearcular region, cells C and 
Sc, stigma, and seams along vein Cu and on anterior cord 
darker; paler longitudinal streaks in cells R, M, and 1st A in 
female sex; veins black. Venation: Rs shorter than cell 1st M2 
and a little longer than m-cu. 

Abdomen with basal segment velvety black; tergites two to 
five, inclusive, orange, narrowly bordered on the lateral and 
caudal portions by black; succeeding tergites and hypopygium 
black; basal sternites conspicuously bicolorous, black, variegated 
more extensively on sides of basal ring with yellow, less ex- 
tensively and distinctly on outer ring just before apex. 

Habitat — Formosa. 

Holotype, male, Hori, altitude 2,340 feet, June 7, 1934 (Gres- 
sitt) . 

Allotopotype, female. Paratopotype, female. 

The nearest ally is LonguHo (Longurio) rubriceps Edwards, 
likewise from Formosa, which differs in the coloration of the 
head, legs, and pattern of the basal abdominal sternites, and in 
the elongate Rs. 

TIPULA (ACUTIPULA) ALBOPLAGIATA sp. nor. Plate U fl». 8; Pl»t« 1, llrf. J6. 1«. 

Allied to Tipula (Acutipula) mnnda; antennal flagellum 
weakly bicolorous, the basal enlargements of the individual seg- 
ments a trifle darker than the remainder; pleura yellow; legs 
black; wings very strongly tinged with brown, variegated by 
a whitish obliterative area before cord and by a pale yellow 
longitudinal streak behind vein Cu; abdomen yellow, the outer 
segments brownish black; male hypopygium with tergal lobe 

190499 8 

342 The Philippine Journal of Science i»85 

bifid; inner dististyle with outer lobe terminating in a very 
slender, acute spine; eighth sternite fringed with yellow setae. 

Male. — ^Length, about 17 millimeters ; wing, 19.5. 

Frontal prolongation of head brown; nasus dark brown, con- 
spicuous; palpi black. Antennse with scape and pedicel dark 
brown ; flagellum very weakly bicolorous, the basal enlargement 
a trifle darker brown than the remainder. Head dark brownish 

Pronotum uniformly reddish brown. Mesonotal praescutum 
gray, with four more olive-gray stripes that are poorly defined 
against the ground, the stripes feebly delimited by very narrow 
darker borders; scutal lobes brownish gray, the median region 
more yellowish; posterior sclerites of mesonotum chiefly yellow 
poUinose. Pleura yellow, the pleurotergite gray pruinose. Hal- 
teres brownish black throughout. Legs with the coxae and 
trochanters yellow; remainder of legs black, excepting only the 
very narrow femoral bases which are yellow. Wings (Plate 1, 
fig. 3) very strongly tinged with brown, the prearcular region 
and cells C and Sc even darker brown; stigma small, yellow; 
ground color variegated by a conspicuous whitish obliterative 
area before the cord, extending from the proximal end of the 
stigma into cell 1st M2 but not or scarcely invading cell M3; a 
conspicuous, pale yellow, longitudinal stripe extending almost the 
entire length of wing immediately behind the basal section of 
vein Cui, involving cell Cui and the adjoining portion of cell 
Cu; veins dark brown. Macrotrichia of veins sparse; squama 
with setae. Venation : Petiole or cell Mi shorter than m. 

Abdomen yellow, the tergites with a narrow, brown, sub- 
lateral border; segments six to nine, inclusive, more uniformly 
brownish black. Male hypopygium (Plate 2, fig. 25) with the 
ninth tergite (Plate 2, fig. 26, 9i) narrowed at outer end into 
a depressed yellow lobe, the apex of which has a deep U-shaped 
median notch, the slender lobes thus formed set with blackened 
spines. Outer dististyle, od, flattened, relatively narrow, the 
outer margin nearly straight, the inner one evenly curved. 
Inner dististyle, id, with the outer lobe terminating in a very 
slender, acute spine; inner lobe darkened, simple, the apex 
obtuse. Eighth sternite (Plate 2, fig. 26, 85) with the caudal 
margin transverse, fringed with yellow setae, those at the outer 
lateral portions a little longer and slightly stouter. 

Habitat. — Formosa. 

Holotype, male, Arisan, altitude 7,640 feet. May 24, 1934 

56,3 Alexander: Tipulidm from Eastern Asia, XXIII 343 

Tipula (Acutipula) alboplagiata is very different from other 
regional members of the munda group in the black legs, darkened 
wings, and structure of the male hypopygium, especially of the 
inner dististyle which is distinct from that of all other de- 
scribed species known to me. 

TIPULA (SCHUMMELIA) INDIFFERENS sp. nov. Plate 1, fig. 4; Plate 2, fi^s. 27, 28. 

Belongs to the continuata group; antennse with flagellar seg- 
ments unicolorous; mesonotal prsescutum obscure brownish 
yellow, with three brownish gray stripes, the median one divided 
by a capillary blackish line; knobs of halteres chiefly light 
yellow; wings grayish brown, the stigma dark brown; restricted 
paler brown clouds in outer end of cell 1st M2, on m-cu, and on 
distal section of Cui ; squama naked ; abdomen orange, segments 
six to nine uniformly black. 

Maie.— Length, 12 to IS millimeters ; wing, 13 to 14.5. 

Frontal prolongation of head dark brown above, paling to 
yellow beneath and on sides ; nasus elongate. Antennse (male) 
of moderate length, if bent backward extending about to mid- 
distance between the bases of the wings and halteres; scape and 
pedicel yellow; basal two or three flagellar segments brownish 
yellow, the succeeding segments passing into black; verticils a 
little longer than the segments. Head brownish gray, more or 
less pruinose. 

Mesonotal prsescutum obscure brownish yellow, with three 
brownish gray stripes, the median one divided by a capillary 
more-blackish line; lateral stripes vaguely bordered by darker, 
especially along their mesal edge ; centers of scutal lobes dark- 
ened; scutellum infuscated basally, margined with yellow; post- 
notum yellow. Pleura chiefly yellow or brownish yellow, the 
pleurotergite darker and weakly pruinose. Halteres dusky, the 
base of stem obscure yellow, the knob chiefly light yellow. Legs 
with the coxsB and trochanters yellow; femora obscure yellow, 
before the tips narrowly blackened, the extreme apices restrict- 
edly pale; tibiae brown, the tips slightly darker; tarsi black. 
Wings (Plate 1, fig. 4) with the ground color grayish brown, 
very weakly patterned with darker; stigma oval, dark brown; 
paler brown clouds across outer end of cell 1st M2, along m-cu 
and outer section of vein Cui; obliterative area extensive, from 
the outer end of cell R into base of cell M3; veins dark brown. 
Squama naked. Venation: R14.2 entire but without trichia ex- 
cept at base; Rs shorter than m-cu; petiole of cell Mi shorter 
than m. 

344 The Philippine Journal of Science i^ss 

Abdomen orange, the pleural region narrowly dusky; seg- 
ments six to nine uniformly black, only the outer dististyle pale 
yellow. Male hypopygium (Plate 2, fig. 27) with the caudal 
border of the tergite (Plate 2, fig. 28, 9t) deeply emarginate, 
blackened, fringed with delicate yellow setae. Outer dististyle, 
od, slightly flattened, narrowed outwardly, the tip obtuse. Inner 
dististyle as shown (Plate 2, fig. 28, id) ; setse of outer margin 
approximately 20 in number. 

Habitat — Kashmir. 

Holotype, male, Srinagar, altitude 5,000 feet, June 9, 1934 
(Miss Hutchinson). Paratopotype, male; paratypes, 3 males, 
Kaj-Nag Range, altitude 8,000 feet. May 30 to June 1, 1934. 

Tipula (Schummelia) indiff evens is very distinct from the 
other regional members of the subgenus in the almost unpat- 
terned wings and the conspicuous coloration of the abdomen. 
The nearest allies would seem to be continuata Brunetti and 
xanthopleura Edwards, both with a very distinct pattern of the 
wings and abdomen. 

TIPULA (SCHUMMELIA) NIGROCELLULA sp. nov. Plate 1, fig. 5; Plate 2, flff. 29. 

Belongs to the continuata group; praescutal stripes distinct, 
narrowly bordered by darker brown; thoracic pleura striped 
longitudinally with brown; knobs of halteres extensively yellow; 
femora with a narrow, nearly terminal, black ring; wings whitish 
subhyaline, with a conspicuous brown pattern ; cell Sc uniformly 
blackened; cell 2d A yellow, variegated by two dark clouds; 
squama naked ; abdominal tergites reddish, narrowly lined with 
darker; a subterminal black ring on abdomen; male hypopygium 
with the caudal margin of tergite tridentate; basistyles bearing 
conspicuous pencils of reddish setse. 

Male, — Length, 12 to 13 millimeters; wing, 12.5. 

Female, — Length, 17 to 19 millimeters ; wing, 15 to 16. 

Frontal prolongation of head yellow above and beneath, nar- 
rowly darkened on sides; nasus elongate, yellow. Antennse 
(male) of moderate length, if bent backward extending approx- 
imately to wing root; scape and pedicel light yellow; flagellar 
segments dark brown, the extreme outer end of the individual 
segments a trifle paler; longest verticils subequal to the seg- 
ments. Head brownish yellow, clearer yellow on anterior ver- 
tex; a narrow brownish median line. 

Mesonotal praescutum yellow, with three conspicuous brown 
stripes that are narrowly bordered by darker brown, the median 
one further divided by a dark vitta ; scutal lobes chiefly covered 

56, 3 Alexander: Tipulidx from Eastern Asia, XXIII 345 

by brown areas, the median region of scutum pale; scutellum 
and mediotergite chiefly darkened medially, paler laterally. 
Pleura yellow, conspicuously striped longitudinally with dark, 
including a dorsal stripe extending from the propleura across 
the ventral anepisternum, and a more interrupted, fainter area 
on the ventral sternopleurite and meron. Halteres with the 
stem obscure yellow, a little brighter at base ; base of knob brown, 
the apex conspicuously light yellow. Legs with the coxse and tro- 
chanters pale yellow; femora brownish yellow, immediately 
before tip with a narrow black ring; tibise brown, paler basally, 
blackened outwardly; tarsi black. Wings (Plate 1, fig. 5) 
whitish to pale yellow subhyaline, handsomely patterned with 
brown; prearcular region and cell C clearer yellow; cell Sc 
uniformly blackened, excepting the yellow extreme outer end; 
stigma dark ; the dark discal pattern includes conspicuous clouds 
across m-cu and fork of Cu; at near midlength of cell M, and 
at outer end of cell 1st Ms, together with the adjoining veins; 
outer radial cells weakly washed with brown; cell 2d A yellow, 
conspicuously variegated by a dark cloud at end and by another 
at midlength of margin; veins dark, except in the obliterative 
area. Squama naked. Venation: R1+2 entire; petiole of cell 
Ml subequal to or shorter than m. 

Abdominal tergites chiefly reddish, the sternites more yellow; 
tergites narrowly and inconspicuously lined medially and later- 
ally with black; a black subterminal ring, involving tergites six 
to nine and sternites seven and eight; remainder of the hypo- 
pygium yellow. Male hypopygium with the tergite (Plate 2, 
fig. 29, 9^ transverse, the emarginate caudal end tridentate, 
with smaller blackened teeth on either side, directed mesad, 
together with a paler median tooth. Region of basistyle, on 
either side, with a pencil of reddish setae, directed ventrad. 
Outer dististyle relatively long and slender, subterete, narrowed 
gradually to the apex. Ninth sternite on either side of ventral 
portion, immediately caudad of the margin of the eighth sternite, 
produced ventrad into a reddish fingerlike lobe. 

Habitat — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 8,000 feet, June 1, 
1934 (Miss Hutchinson). Allotopotype, female. Paratopo- 
types, 1 male, May 25, 1934; 1 male, 1 female, altitude 9,000 
feet, May 30 to June 3, 1934. 

The present species is readily told from all other regional 
members of the subgenus by the uniformly blackened subcostal 
cell and by the rather peculiar structure of the male hypopygium. 

346 The Philippine Journal of Science 1935 


Female. — Length, about 17 millimeters; wing, 15.5. 

Generally similar to T. (S.) nigrocellula sp. nov., differing in 
several important details. 

Nasus short and stout. Antennae with scape brownish yellow. 
Squama with setse. Wings with cell Sc light yellow, concolorous 
with cell C ; outer two-thirds of cell 1st M2 abruptly darkened ; 
cell Ml conspicuously pale; dark seam along m-cu scarcely in- 
vading cell M4 ; cell 2d A uniformly suffused with pale brown. 

Compared with T. (S.) continuata Brunetti, which likewise 
has a group of setae on the squama, the present fly differs con- 
spicuously in the coloration of the antennae and body, and in 
the venation and wing pattern, especially the much longer Rs, 
which is somewhat longer than m-cu and nearly twice as long 
a-s R 2+3; elongate cell 1st M2, with the second section of vein 
Mi^2 nearly two-thirds as long as vein Mi; and in the broader 
cell M4. 

Habitat. — Kashmir. 

Holotype, female, Kaj-Nag Range, altitude 8,000 feet, June 1, 
1934 {Miss Hutchinson) . 

TIPULA (LUNATIPULA) TRIALBOSIGNATA sp. nov. Plate 1, fig. 6; Plate 2, figs. 
30, 31. 

Antennae bicolorous; mesonotal praescutum with ground color 
yellow pollinose, narrowly or obsoletely striped with darker 
yellow; wings grayish, variegated with darker gray and whit- 
ish areas, including a broad, incomplete, poststigmal fascia; a 
dark spot in cell Cu, preceded and followed by whitish areas; 
abdomen yellow, in the female with the tergites narrowly tri- 
lineate with darker yellow; male hypopygium with the basistyle 
produced caudad into a subobtuse chitinized point; eighth ster- 
nite with caudal brush of long yellow setae. 

Male. — Length, about 15 millimeters; wing, 15.5. 

Female. — Length, about 20 millimeters ; wing, 18.5. 

Frontal prolongation of head brownish yellow above, darker 
laterally; nasus distinct. Antennae (male) of moderate length, 
if bent backward extending to the wing root or shortly beyond ; 
scape and pedicel yellow; flagellar segments bicolorous, brownish 
black basally, the outer end yellow, narrowly so on the first 
flagellar segment; outer flagellar segments more uniformly 
darkened. Head brownish yellow. 

Mesonotal praescutum with the ground color yellow pollinose, 
with four narrow brown stripes that represent the lateral stripes 
and posterior outer borders of the usual median stripe ; in male. 

5e,8 Alexander: Tipulidse from Eastern Asia, XXIII 347 

prsescutal stripes obsolete or nearly so; scutal lobes more or less 
variegated with darker color; scutellum and mediotergite yellow. 
Halteres yellow, the knobs infuscated. Legs with the coxse and 
trochanters yellow; femora obscure yellow to brownish yellow, 
the tips narrowly dark brown, preceded by a slightly clearer, 
more yellowish ring; tibise and basitarsi obscure yellow, the tips 
narrowly dark brown; outer tarsal segments dark brown. 
Wings (Plate 1, fig. 6) with the ground color grayish, variegated 
by pale and darker areas ; the white markings include one before 
stigma; a broad poststigmal fascia, extending from margin to 
cell 1st M2, and an obliterative streak across base of cell 1st 
M2, extending from cell R into bases of cells M3 and M4, usually 
disconnected from the poststigmal fascia but, in cases, confluent; 
additional white areas on either side of a dark cloud in cell Cu ; 
the dark areas include the stigma and a confluent area along 
cord, together with the dark spot in cell Cu above described; 
narrow and less evident dark seams on posterior cord and outer 
end of cell 1st M2; veins brown to brownish yellow. Macro- 
trichia of veins small and relatively sparse ; squama with sparse 
setse. Venation: R^.^^ entire; petiole of cell Mi shorter than m. 

Abdomen entirely yellow or orange-yellow; in female, with 
indications of three narrow brown tergal stripes. Male hypo- 
pygium (Plate 2, fig. 30) with the tergite chiefly separated from 
the sternite. Ninth tergite (Plate 2, fig. 31, 9^) extensive, the 
caudal margin terminating in two small acute points that are 
separated from one another by a very shallow emargination ; as 
in Vestiplex, these points are the dorsal manifestation of a tergal 
saucer that has swung caudad and ventrad so as to lie on the 
lower surface of the tergite. Basistyle, b, entire, the outer end 
produced into a subobtuse chitinized point ; ventral portion with 
an oval lobe set with long yellow setse. Outer dististyle slender, 
pale, from a slightly enlarged base. Inner dististyle, id, with 
the outer margin heavily blackened and corrugated. Eighth 
sternite, Ss, with a dense brush of long yellow setse but without 
lobes or other armature. Ovipositor with the cerci rather stout, 

Habitat. — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 9,000 feet, June 3, 
1934 (Miss Hutchinson). Allotopotype, female. Paratopo- 
types, 1 male, 1 female. 

The darkened cloud in cell Cu is suggestive of the condition 
found in many species of Acutipula and other subgeneric groups 
of TipvXa, but the present fly seems correctly referrable to Lunor- 

348 2"^^ Philippine Journal of Science i936 

tipula. The wing pattern and the structure of the male hypo- 
pygium, especially the basistyle, readily separate the species 
from other described Himalayan members of the genus. 

TIPULA (OREOMYZA) ARISANENSIS Edwards. Plate 2, flfirs. 32, 33. 

Tipula arisanensis Edwards, Ann. & Mag. Nat. Hist. IX 8 (1921) 

The types, two females, were from Arisan, Formosa, collected 
May 24, 1917, by Shiraki. One male, from Pianan-ambu, For- 
mosa, May 11, 1932, collected by Gressitt, seems undoubtedly to 
belong to this species and is described herewith as allotype. 

The male hypopygium is of rather remarkable construction, 
much as in Tipula foliacea Alexander, yet with the details quite 

Ninth tergite (Plate 2, fig. 32) very large and massive, fused 
with the sternite on basal half, the outer half indicated by a 
gently curved suture ; lateral tergal arms produced into powerful 
flattened blades that extend caudad about as far as the level 
of the tips of the inner dististyle ; median area of tergite with 
two low, triangular points. Viewed from the side (Plate 2, 
fig. 33) the tergal blades are seen to be strongly decurved, nar- 
rowed outwardly, with a shallow notch close to tip. 

Male. — Length, about 9.5 millimeters; wing, 11.5. 

Allotype, male, Pianan-ambu, Formosa, May 11, 1932 (Gres- 

In T. foliacea the tergal blades are much longer, extending far 
beyond the level of the dististyles; the intermediate lobes of 
the tergite are obtusely rounded. 

TIPULA GRESSITTI sp. nor. Plate 1, figr. 7; Plate 2, fig. 34. 

General coloration gray, the prsescutum with four dark brown- 
ish gray stripes; antennae (male) relatively elongate, bicolorous; 
front silvery white; tips of femora broadly black; wings gray, 
variegated by yellow and darker areas ; male hypopygium very 
small and of simple structure; ovipositor with short, fleshy 

Male. — Length, 9 to 10 millimeters ; wing, 13 ; antenna, 4. 

Female. — ^Length, about 9 to 10 millimeters; wing, 11. 

Frontal prolongation of head dark brown, sparsely pruinose ; 
nasus long and conspicuous, tufted with black setae ; palpi black. 
Antennae (male) elongate, the thirteenth segment microscopic; 
antenna approximately as long as the combined head and thorax ; 
scape and pedicel yellow; first flagellar segment yellow, weakly 
darkened at base; succeeding segments conspicuously bicolorous. 

56,3 Alexander: Tipulidx from Eastern A^ia, XXIII 349 

the basal enlargement black, the remainder yellow; outer seg- 
ments more obscure but all except the twelfth retaining the 
bicolorous appearance; verticils a little shorter than the seg- 
ments. Head with the front silvery white ; anterior vertex and 
orbits yellow; disk of vertex more infumed, variegated by an 
elongate brown area on either side of the median furrow, the 
inner ends of the marks dilated, the narrow outer portion paral- 
leling the orbits; an additional smaller dark spot immediately 
behind each antennal fossa. 

Mesonotum gray, the prsescutum with four dark brownish 
gray stripes, the intermediate pair separated only by a capillary 
line ; scutum dark brownish gray, the anterior portion darkened 
by a posterior prolongation of the lateral prsescutal stripes; 
posterior sclerites of mesonotum blackened, sparsely pruinose. 
Pleura black, heavily pruinose, more heavily so on ventral ster- 
nopleurite; dorsopleural membrane obscure yellow. Halteres 
chiefly pale yellow, darkened only by abundant black setse near 
base of club. Legs with coxae brownish gray, paler apically; 
trochanters yellow; femora obscure brownish yellow, the tips 
broadly and conspicuously blackened; tibiae and tarsi brownish 
black to black; tibial spur formula apparently 1-1-2. Wings 
(Plate 1, fig. 7) with the ground color gray, variegated by 
brown and yellow areas; prearcular region and cell Sc yellow; 
cell C uniformly infuscated; stigma dark brown; a distinct 
brown cloud on anterior cord; the yellow color includes areas 
before and beyond stigma; near base and near outer end of cell 
M ; near outer ends of cells Mi to M4, inclusive ; most of cell Cu 
and areas in cell 1st A; cell 2d A more uniformly darkened; 
veins brown. Squama naked; macrotrichia of veins relatively 
long and numerous. Venation: R14.2 entire; petiole of cell Mi 
variable in length, in the allotype nearly lacking so that cell Mi 
is subsessile; cell 2d A relatively narrow. 

Abdomen brown, the basal two tergites variegated with ob- 
scure yellow; lateral margin of second tergite blackened; hypo- 
pygium dark. Male hypopygium (Plate 2, fig. 34) very small 
and of simple structure, among the most generalized so far dis- 
covered in the genus. Ninth tergite separated from the fused 
sternite-basistyle ; suture between basistyle and sternite indi- 
cated beneath, the former region extensive. Ninth tergite, 9t, 
transverse, the caudal margin with a broad shallow emargina- 
tion, the lateral lobes obtuse, with abundant normal setse. Outer 
dististyle, od, broadly flattened, the apex obtuse. Inner disti- 

350 The Philippine Journal of Science 1935 

style, id, simple, the apical beak very long and slender. Eighth 
sternite simple, weakly sheathing, the outer end membranous but 
with abundant setse. Ovipositor with short fleshy valves. 

Habitat. — Formosa. 

Holotype, male, Arisan, altitude 7,640 feet, May 24, 1934 
(Gressitt). AUotopotype, female. Paratopotypes, 2 males, 1 

Tipula gressitti is named in honor of the collector, who has 
secured many interesting Tipulidse in the high mountains of 
Formosa. I cannot place the fly in any of the now numerous 
subgenera of the vast genus Tipula. The Eastern Nearctic 
Nobilotipula Alexander has an ovipositor of somewhat similar 
structure, but the present fly scarcely seems to be consubgeneric. 
In its general appearance, especially in body coloration and wing 
coloration, the species bears a considerable resemblance to sev- 
eral other regional members of the subgenera Acutipula, Luna- 
tipula, Oreomyza, and Vestiplex, but in the details of structure 
is very distinct. 

NEPHROTOMA ATROLATERA sp. nov. Plate 1, flgr. 8. 

Mesonotal prsescutum with the central portion largely covered 
by a shield-shaped pearly area that is margined in front and on 
sides by a velvety-black border ; scutellum and mediotergite light 
yellow ; pleura yellow ; abdominal tergites yellow, trilineate with 
dark brown; sternites uniformly orange-yellow. 

Female. — ^Length, about 14 millimeters; wing, 13.5. 

Frontal prolongation of head light sulphur yellow ; nasus with 
long black setse; mouth parts more fulvous; palpi dark brown. 
Antennae with basal segments brownish yellow, the flagellum 
passing into darker brown. Front light sulphur yellow; pos- 
terior portion of head dark orange, more yellowish behind but 
without evident occipital brand. 

Pronotum entirely light sulphur yellow. Mesonotal prsescu- 
tum with three confluent pearly stripes, the shield-shaped area 
thus formed bordered by velvety black, including a narrow cepha- 
lic border, broader lateral margins, and a deep triangular inva- 
sion of the humeral field; scutum with centers of lobes pearly, 
narrowly but completely bordered by velvety black; median re- 
gion of scutum yellow; scutellum and mediotergite light sulphur 
yellow, with a narrow transverse dark line separating the two ; 
mediotergite with numerous short black setae. Anterior lateral 
pretergites sulphur yellow. Pleura yellow. Halteres dusky, the 
knobs yellow. Legs with the coxae and trochanters light yellow ; 
femora and tibiae light brown, the tips weakly darkened; tarsi 


Alexander: Tipulidae from Eastern Asia, XXIII 351 

black. Wings (Plate 1, %. 8) somewhat teneral in the unique 
type, subhyahne; stigma and a seam along the cord narrowly 
darker; veins brown. Venation: Cell M, broadly sessile; m-cu 
uniting with M just before departure of vein M4 

Abdominal tergites yellow, conspicuously trilineate with dark 
brown, the stripes entire or virtually so; sternites orange-yellow, 
the caudal borders of the segments restrictedly paler yellow. 

Habitat, — Formosa. 

Holotype, female, Hassensan, altitude 4,875 feet, June 22 
1934 (Gressitt). ' 

The present fly is amply distinct from other described species 
of the genus in the pattern of the mesonotal pr^scutum and 
scutum. The presence of short, dense, black set^ on the post- 
notal mediotergite reminds one of Nephrotoma medipubera 
Edwards (eastern Java), which is otherwise a very different fly. 

DOLICHOPEZA (NESOPEZA) LUGUBRIVESTIS «p. nov. Plate 1, fi^. 9; Plate 2, ng. 35. 

General coloration dark brown, the pr^scutum with three 
more-glabrous, somewhat nacreous stripes; pleura dark brown; 
legs black, only the outer tarsal segments restrictedly yellowish 
white; wings with the ground color strongly blackened; stigma 
darker, preceded and followed by whitish areas; Rs short, a 
little less than m-cu; male hypopygium with the lateral lobes 
of the tergal plate broadly rounded; gonapophyses long and 
conspicuous, the tips acute. 

MoZ^.— Length, about 12 millimeters; wing, 13. 

Frontal prolongation of head brown ; palpi black. Antennge of 
moderate length, if bent backward extending approximately to 
base of abdomen; scape and pedicel brownish yellow, flagellum 
black; verticils abundant, the longest on the upper face and well 
scattered over the length of the segment. Front and vertical 
tubercle brownish yellow, the posterior portions of head more 
uniformly dark brown. 

Mesonotal prsescutum dull dark brown, with three more-gla- 
brous, nacreous stripes, the median one broad and well defined, 
the lateral pair less clearly delimited; posterior sclerites of 
notum dark brown, the median area of scutum and mediotergite 
paler. Pleura dark brown, the dorsopleural region paler. Hal- 
teres with the stem dusky, narrowly obscure yellow at base, the 
knobs dark brown. Legs with the coxae dark brown, paler api- 
cally; trochanters obscure yellow; femora and tibiae black, the 
former restrictedly paler at base; tarsi black basally, the outer 
segments narrowly and restrictedly yellowish white, more ex- 

352 The Philippine Journal of Science 1935 

tensively so on the posterior legs where the brightening involves 
about the outer two-thirds of the second tarsal segment and 
the succeeding segments; on the anterior legs the bright color 
of the tarsi is even more obscured. Wings (Plate 1, fig. 9) with 
the ground color strongly blackened, the stigma even darker, 
preceded and followed by whitish areas; a restricted darkened 
cloud on anterior cord; veins brownish black. Venation: Rs 
short, a little less than m-cu ; medial forks of moderate depth. 

Abdominal tergites dark brown, the incisures restrictedly 
more blackened, outer segments, including hypopygium, more 
uniformly blackened. Male hypopygium (Plate 2, fig. 35) with 
the lateral lobes of the tergal plate, 9t, broad, their caudal 
margins broadly rounded, i^deagus and subtending apophyses, 
g, very long and conspicuous, as illustrated ; apophyses at base 
with a small cylindrical lobe but beyond this point simple, the 
tips acute. 

Habitat — Formosa. 

Holotype, male, Arisan, altitude 7,475 feet, May 25, 1934 
(Gressitt) . 

Dolichopeza (Nesopeza) lugubrivestis is most nearly allied to 
D. (N.) idiophallus (Alexander) and D. (N.) tarsalba (Alexan- 
der) , differing from both in the coloration of the body and wings, 
the venation, especially the short Rs, and the details of structure 
of the male hypopygium. 


LIMONIA (LIMONIA) SUBHOSTILIS sp. nor. Plate 1, tig, 10; Plate S, fig, 38. 

General coloration brown, the prasscutal stripes not or scarcely 
indicated; antennae black throughout; each of the outer flagellar 
segments with a single very long verticil, these unilaterally 
arranged; outer flagellar segments gradually increasing in 
length to the last; halteres elongate; legs brown, the femoral 
tips not brightened; wings broad, suffused with gray; stigma 
lacking; Scg and R14.2 both very long; male hypopygium with a 
single dististyle, this terminating in a slender apical beak. 

Male. — ^Length, about 10 millimeters; wing, 11.8. 

Rostrum brown; palpi dark brown. Antennae dark brown 
throughout; outer flagellar segments with a single verticil of 
unusual length, placed on outer face near base, this seta ap- 
proximately twice the length of the segment bearing it ; remain- 

56, 8 AlexaTider: Tipididse from Eastern Ada, XXIII 353 

ing verticils relatively small and insignificant; flagellar seg- 
ments beyond midlength of organ becoming progressively more 
elongate, the terminal segment longest. Head dark brown. 

Mesonotum almost uniformly medium brown, the surface 
polished; praescutal stripes not or scarcely defined, the lateral 
pair indicated chiefly by a slightly darkened lateral portion. 
Pleura testaceous-brown. Halteres unusually long and slender, 
dark brown. Legs with the coxae and trochanters brown; re- 
mainder of legs medium brown, the femoral bases not or scarce- 
ly brightened; claws with a long slender spine at near mid- 
length, with a second smaller, more basal tooth. Wings 
(Plate 1, fig. 10) broad, almost uniformly suffused with gray; 
stigma not or scarcely indicated; veins brown. Macrotrichia 
of veins relatively long and conspicuous, on all longitudinal veins 
beyond cord, extreme distal end of vein 1st A and outer two- 
thirds of 2d A. Venation: Scg very long, as in hostilis, several 
times as long as Sci and considerably longer than R2; R1+2 
elongate ; cell 1st M2 wide ; m-cu close to fork of M. 

Abdomen dark brown; hypopygium somewhat paler. Male 
hypopygium (Plate 3, fig. 36) with the tergite, 9t, narrowed 
outwardly, the apex abruptly thin and glabrous, with a U- 
shaped median notch, the lateral lobes thus formed low and 
subtruncate. Dististyle, d, single, its apical beaklike portion 
unusually slender. Gonapophyses, g, with the mesal-apical lobe 
broad, ^deagus broad at apex. 

Habitat. — Formosa. 

Holotype, male, Arisan, altitude 7,312 feet. May 27, 1934 
(Gressitt) . 

The nearest described ally of the present fly is undoubtedly 
Limonia (Limonia) kostilis Alexander (Szechwan-Tibet border, 
at high altitudes), which differs conspicuously in the abruptly 
yellow tips of the femora and in slight details of venation and 
body coloration. 

LIMONIA (DICRANOMYIA) TRANSFUGA sp. nov. Plate 1, fig. 11. 

Mesonotal prsescutum with a broad median brown stripe, the 
lateral borders of the sclerite broadly golden yellow; pleura 
pruinose; knobs of halteres dark brown; femora blackened at 
tips, the fore femora more extensively so; wings tinged with 
yellow, sparsely patterned with brown; abdomen with basal six 
segments orange-yellow, the remaining outer segments black. 

354 The Philippine Journal of Science 1935 

Female. — ^Length, about 9.5 millimeters; wing, 10. 

Rostrum brown, sparsely pruinose; palpi black. Antennae 
black throughout; basal jfiagellar segments subglobular, the outer 
segments oval to elongate-oval. Head brownish gray. 

Mesonotum with a broad median brown stripe, the lateral 
portions of the sclerite broadly and conspicuously golden-yellow 
pollinose; lateral praescutal stripes scarcely indicated; scutal 
lobes variegated with darker yellow; mediotergite black, 
sparsely pruinose, the lateral portions paling to yellow. Pleura 
yellow, the mesopleura and pleurotergite heavily pruinose. 
Halteres with basal portion of stem yellow, the outer portion 
and knob dark brown. Legs with the coxse and trochanters 
yellow; fore femora black, the basal third yellow; remaining 
femora yellow, the tips more narrowly blackened, not including 
more than the distal fourth or fifth ; tibiae and tarsi dark brown 
to brownish black. Wings (Plate 1, fig. 11) with a strong 
yellow tinge, the costal border and outer radial field more- 
saturated brownish yellow; stigma and narrow seams along 
cord and outer end of cell 1st M2 dark brown ; a brownish seam 
along vein Cu in cell M; veins brownish yellow, darker in the 
clouded areas. Venation: Sci ending a short distance beyond 
origin of Rs, Sc2 some distance before this origin; free tip of 
Scs and R2 in transverse alignment; m-cu close to fork of M, 
subequal to or a little longer than distal section of Cui; anal 
veins at origin subparallel or very slightly convergent. 

Abdomen with the basal six segments orange-yellow, the re- 
maining segments black. Ovipositor with the cerci relatively 
short and unusually slender, gently upcurved to the acute tips, 
a little longer than the hypovalvse. 

Habitat. — Kashmir. 

Holotype, female, Kaj-Nag Range, altitude 9,000 feet, May 
30, 1934 (Miss Hutchinson). 

Limonia (Dicranomyia) transfuga is quite distinct from other 
regional members of the subgenus, superficially bearing a certain 
resemblance to L. (D.) baileyi (Edwards), of Tibet, yet entirely 

DICRANOTA (EHAPHmOLABIS) SUBSORDmA sp. nov. Plate 1, fig. 12; Plate S, Hg, S7. 

General coloration gray, the prsescutum with a darker median 
stripe; antennae black throughout; knobs of halteres infuscated; 
legs brownish black; wings grayish subhyaline, the oval stigma 
brown; R2+3+4 approximately twice the basal section of R5; 
abdominal tergites dark brown, hypopygium brownish yellow; 

66, 3 Alexander: Tipulidae from Eastern Asia, XXIII 355 

male hypopygium with the caudal margin of tergite transverse, 
each outer angle produced into a low setiferous lobe; interbase 
widely expanded on proximal portion. 

Male. — Length, about 5 millimeters ; wing, 6.2. 

Rostrum gray; palpi black. Antennse short, black through- 
out, the scape a little pruinose. Head gray. 

Mesonotum gray, the prsescutum with a darker, plumbeous 
gray, median stripe. Pleura gray. Halteres pale, the knobs 
infuscated. Legs with the coxse pruinose on outer face, the 
inner face pale; trochanters yellow; femora brownish black, the 
bases restrictedly pale; tibise and tarsi brownish black. Wings 
(Plate 1, fig. 12) grayish subhyaline; stigma oval, brown; veins 
dark brown. Venation: R2+3+4 approximately twice the basal 
section of R5; R2 transverse; cell Mi unusually small, the inclos- 
ing branches divergent; m-cu about one-half its length beyond 
fork of M. 

Abdominal tergites dark brown; sternites somewhat paler, 
slightly pruinose; hypopygium brownish yellow. Male hypopy- 
gium (Plate 3, fig. 37) with the tergite, 9i, transverse across 
caudal margin, each lateral portion produced into a low, rounded, 
setiferous lobe. Dististyle, rf, with the outer blade elongate, 
flattened, a little expanded on outer end and here provided with 
several set^. Interbase, i, a yellow blade, very widely expanded 
on proximal portion, gradually narrowed outwardly, the tip 
narrowly obtuse. 

Habitat, — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 8,000 feet, May 18, 
1934 {Miss Hutchinson) . 

The species is most readily told from Dicranota (Rhaphido- 
labis) sordida (Brunetti) by the black antennse and almost uni- 
formly darkened legs. 

DICRANOTA (RHAPHmOLABIS) UNINEBULOSA sp. nov. Plate 1. fig. 13; Plate 8, 
flfiT. 38. 

General coloration gray, the prsescutum without evident 
stripes ; antennal flagellum black ; knobs of halteres weakly dark- 
ened ; femora brownish yellow, the tips weakly darkened ; wings 
subhyaline, stigma brown ; a small brown cloud on anterior cord ; 
R24-8+4 fully one-half longer than basal section of vein R5; 
abdomen brown, sparsely pruinose, the extreme caudal margins 
of the segments pale; hypopygium yellow; male hypopygium 
with the interbase curved, the apex flattened and having the 
margin microscopically serrulate. 

Male. — Length, about 6.5 millimeters; wing, 7.5. 

356 The Philippine Journal of Science 1935 

Rostrum gray; palpi dark. Antennse short; scape and pedicel 
brownish black, flagellum black; antennae 15-segmented ; fla- 
gellar segments short-oval to subglobular. Head dark gray. 

Mesonotum gray, the prsescutum without evident stripes. 
Pleura gray. Halteres pale, the knobs weakly darkened. Legs 
with the coxae pale, gray pruinose ; femora brownish yellow, the 
tips weakly darkened ; tibiae pale brown, the tips, together with 
the tarsi, more brownish black. Wings (Plate 1, fig. 13) sub- 
hyaline ; stigma brown ; a small but distinct brown cloud on ante- 
rior cord ; veins brown. Venation : R2 + 3+4 fully one-half longer 
than the basal section of vein R5; R2 transverse or nearly so; 
M almost in direct alignment with Mi^^f the basal section of 
the latter lacking or virtually so ; m-cu about one-half its length 
beyond fork of M. 

Abdomen brown, sparsely pruinose; extreme caudal margins 
of segments pale; hypopygium yellow. Male hypopygium with 
the inter base (Plate 3, fig. 38, i) curved, the apex flattened, its 
margin microscopically serrulate. 

Habitat — ^Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 8,000 feet. May 26, 
1934 (Miss Hutchinson) . 

Dicranota (Rhaphidolabis) uninehulosa is readily told from 
the other Himalayan species of the subgenus by the large size 
and the pattern of the wings. 

DICRANOTA (RHAPHIDOLABIS) PALLIDITHORAX sp. nov. Plate 1, figr. 14; Plate 
3, fig. 30. 

General coloration of thorax pale brownish yellow; antennse 
dark brown, the scape brownish yellow; knobs of halteres dark- 
ened; wings subhyaline, the stigmal area faintly darker; R24.8+4 
about one-half longer than the basal section of vein R5; R2 
oblique; male hypopygium with the interbase on mesal face at 
near midlength produced into a spinous point. 

Male. — Length, about 7 millimeters; wing, 7.8. 

Ferrmle. — ^Length, about 8 millimeters; wing, 8.5. 

Rostrum brown, paler laterally ; palpi brown. Antennse short, 
15-segmented; scape brownish yellow, remaining segments dark 
brown. Head light gray throughout. 

Pronotum and mesonotum, together with the pleura, entirely 
pale brownish yellow to yellow, without markings. Halteres 
pale, the knobs darkened. Legs with the coxs& and trochanters 
yellow; femora brownish yellow, clearer yellow basally, darker 
outwardly; tibiae and tarsi dark brown. Wings (Plate 1, fig. 

56, 3 Alexander: Tipulidse from Eastern Asia, XXIII 357 

14) subhyaline, iridescent; stigmal area faintly darker, pale 
brown ; veins brown. Venation : Rs of moderate length ; R2+3+4 
about one-half longer than the basal section of vein R5; R2 
oblique, exceeding one-half R^+g; ni-cu about its own length 
beyond the fork of M. 

Abdominal tergites dark brown, the caudal borders of the 
segments very narrowly pale ; basal sternites obscure yellow, the 
outer segments somewhat more darkened; hypopygium chiefly 
yellow. Male hypopygium with the interbase (Plate 3, fig. 39, 
i) of characteristic shape, on mesal face at near midlength 
produced into an acute spinous point, the apical portion slender, 
narrowed before the oval distal end. 

Habitat, — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 8,000 feet, May 22, 
1934 (Miss Hutchinson) . Allotopotype, female. 

Dicranota (Rhaphidolabis) pallidithorax is very distinct from 
the other Asiatic species of the subgenus so far described in 
the major size, the pale coloration of the thorax, and the 
structure of the interbase of the male hypopygium. 

DICRANOTA (AMALOPINA) FUMICOSTATA sp. nov. Plate 1. fig. 15; Plate 3, fig. 40. 

Mesonotal prsescutum yellow, with a median darker stripe; 
posterior sclerites of mesonotum darkened; legs yellow; wings 
yellow, the costal border to apex broadly infumed, including all 
of cells C and Sc ; smaller darker areas on certain of the veins 
and crossveins; no supernumerary crossvein in cell Ri; R2+3-f4 
short ; cell 1st M2 closed ; male hypopygium with both the tergal 
arms and the interbases appearing as stout flattened blades that 
terminate in acute spines. 

Male, — Length, about 5.5 millimeters; wing, 6.5. 

Rostrum and palpi brownish black. Antennae with the scape 
brownish black; pedicel brownish yellow; flagellum light yellow, 
only the outer two or three segments more darkened. Head 
brown ; eyes relatively large and protuberant. 

Cervical sclerites and central portion of pronotum dark brown. 
Mesonotal prsescutum yellow, with a more brownish median 
stripe; lateral stripes narrow and scarcely evident; posterior 
sclerites of mesonotum darker brown, this color including the 
posterior border of the pleurotergite. Remainder of pleuroter- 
gite and all of the pleura pale yellow. Halteres with the stem 
yellow, its outer end and the knob infuscated. Legs with the 
coxae and trochanters pale yellow; remainder of legs light yellow, 

290499 ^9 

358 The Philippine Journal of Science isss 

only the last tarsal segment darkened. Wings (Plate 1, fig. 
15) with the ground color yellowish, the costal border to apex 
broadly infumed, the darkening including all of cells C and Sc, 
together with the basal third of cell R ; restricted darker brown 
areas at origin of Rs, along cord, R2, outer end of cell 1st M2 
and fork of M1+2 ; veins pale, darker in the infumed areas. 
Venation: Sc2 lying at near three-fifths the distance between 
arculus and origin of Rs; no supernumerary crossvein in cell 
Ri; cell R3 short-petiolate, R2+3+4 being represented by an ele- 
ment that is subequal to or shorter than the basal section of 
Rs; cell 1st Mg closed; m-cu about one-half its length beyond 
the fork of M. 

Abdominal tergites brown, the sternites paler; abdomen 
darker apically. Male hypopygium (Plate 3, fig. 40) with the 
lateral arms of the tergite, 9t, appearing as powerful erect rods, 
the tip incurved to a short acute spine; median area of tergite 
slightly convex, with abundant setae. Interbases, i, large and 
powerful, appearing as sinuous yellow blades, the tip of each 
narrowed into an acute spine. Dististyle unusually simple, the 
outer lobe with the usual close-set spines, the inner blade with 
numerous setae. 

Habitat. — Formosa (north). 

Holotype, male, Urai, altitude 1,500 feet, April 1, 1932 (Gres- 
sitt) . 

Dicranota (Amatopina) fumicostata is very different from 
the other regional species of the subgenus, the distinctions being 
best shown by the accompanying key to the five species now 
known from the Japanese Empire. 

Key to the Japanese species of Amatopina, 

1. Cell 1st Ma open by atrophy of m; fore and middle femora brownish 

black, the posterior femora and remainder of all legs light yellow 

(gibbera and races) 2. 

Cell 1st M2 closed; legs yellow 3, 

2. Wings (male) broad, widest opposite end of vein 2d A. (Japan.) 

gibbera gibbera Alexander. 

Wings (male) narrow, of approximately equal width along the central 

third of length. (Japan.) gibbera karafutonis Alexander. 

3. No supernumerary crossvein in cell Ri; outer ends of radial cells uni- 

formly infumed. (Formosa.) fumicostata sp. nov. 

A supernumerary crossvein in cell Ri; outer ends of radial cells clear 
or with darkenings at ends of veins only 4. 

4. Cells C and Sc undarkened. (Japan.) dicranotoides Alexander. 

Cell C, and usually Sc also, strongly darkened on basal half 6. 

56,8 Alexander: Tipulidx from Eastern Asia, XXIII 359 

6. Outer ends of all longitudinal veins with brown spots and dots. (Si- 
beria and Korea.) sibenca Alexander. 

Outer ends of longitudinal veins undarkened. (Formosa.) 

delectata Alexander. 
It may be noted that all of the above species (excepting sibe- 
rica, a male of which is not available to me at this time) show 
marked distinctions in the structure of the male hypopygium. 

GONOMYIA (GONOMYIA) JUSTA ep. nov. Plate 1, jftff. 16; Plate 3, &g. 41. 

Belongs to the tenella (subcinerea) group; allied to G. (G.) 
nebulicola; antennae black throughout; mesonotum grayish 
brown, restrictedly variegated with yellow; pleura sulphur 
yellow, variegated by reddish brown; legs dark brown; R2+3-H 
strongly arcuated; male hypopygium with the outer lobe of 
basistyle relatively short, only a little longer than the dististyle, 
the latter broadly flattened, bispinous; blackened appendage of 
phallosome not spinous. 

Male, — Length, about 4.2 millimeters ; wing, 5. 

Rostrum yellow; palpi brownish black. Antennae black 
throughout; outer flagellar segments very slender, almost seta- 
ceous, the basal ones more enlarged. Head gray. 

Pronotum and anterior pretergites light yellow. Mesonotum 
uniformly dark grayish brown, the humeral region and sides of 
prsescutum restrictedly yellow; median area of scutum obscure 
yellow; posterior lateral portions of scutal lobes and the broad 
posterior border of scutellum dark yellow. Pleura light sul- 
phur yellow, variegated by reddish or reddish brown on anepi- 
sternum and more extensively on ventral sternopleurite and 
meron. Halteres dusky, the base of stem light yellow. Legs 
with the coxse reddish or yellow, the fore pair darker; trochan- 
ters obscure yellow; remainder of legs dark brown. Wings 
(Plate 1, fig. 16) uniformly tinged with gray, the stigma pale 
brown; veins dark brown. Venation: Sci ending shortly be- 
yond origin of Rs, SC2 opposite this origin; R 2+3+4 strongly 
arcuated ; basal section of R5 reduced; cell 1st M2 relatively small ; 
m-cu at fork of M. 

Abdominal tergites dark brown, the sternites and hypopygium 
more brownish yellow. Male hypopygium (Plate 3, fig. 41) 
with the outer lobe of basistyle, b, relatively short, only a little 
longer than the dististyle. Dististyle, d, broadly flattened, bear- 
ing two unequal spines on outer margin, the outermost long and 

360 The Philippine Journal of Science 1935 

gently curved, with a single seta at base; second spine in axil 
of the first; no carina connecting the inner spine with apex 
of style, as is the case in nebulicola. Phallosome, p, with a 
single blackened appendage, this subobtuse at apex, not spinous. 

Habitat. — Java. 

Holotype, male, Mount Malabar, altitude about 4,000 feet. 
May 26, 1929 (Mackerras). Paratopotype, male. 

Type in the National Collection, Federal Capital Territory, 

The nearest described ally of the present fly is undoubtedly 
Gonomyia (Gonomyia) nebulicola Alexander (Mindanao), which 
differs chiefly in the structure of the dististyle and phallosome 
of the male hypopygium, as contrasted above. 

GONOMYIA (LIPOPHLEPS) TORAJA sp. nov. Plate 1, fig. 17; Plate 3, figr. 42. 

Mesonotum dark gray, the scutellum obscure yellow, pruinose ; 
thoracic pleura with a silvery white longitudinal stripe; knobs 
of halteres yellow; femora obscure yellow, the tips broadly black- 
ened; wings with a grayish tinge, the ground color vaguely 
brightened by paler areas ; veins along cord darkened ; basal sec- 
tion of Rs long ; abdomen dark brown ; male hypopygium with 
two dististyles, the outer a long simple black rod; inner style 
profoundly bifid, its outer arm fusiform, clothed with long con- 
spicuous setse. 

Male. — ^Length, about 3 millimeters; wing, 3.2. 

Rostrum and palpi black. Antennse black; flagellar segments 
(male) with very long verticils. Head brownish yellow, the 
center of vertex darker. 

Sides of pronotum and the anterior lateral pretergites whitish, 
the posterior pretergites more silvery. Mesonotal prsescutum 
dark gray; pseudosutural fovese black; scutum dark gray, in- 
cluding the median area ; scutellum obscure yellow, sparsely prui- 
nose, the base darkened; mediotergite dark, heavily pruinose. 
Pleura dark, including the dorsopleural membrane; a conspic- 
uous silvery longitudinal stripe extending from behind the fore 
coxse to the base of abdomen ; indications of a second, much less 
distinct, obscure yellow stripe above the first, involving the 
pteropleurite, pleurotergite, and cephalic-lateral portions of the 
mediotergite. Halteres dusky, the knobs yellow apically. Legs 
with the coxse dark brown, the fore coxse pale on outer face; 
trochanters brownish yellow; femora obscure yellow, the tips 
broadly black (certain of the femoral tips are broken and it 
cannot be affirmed as to whether or not certain of these areas 

56. 3 Alexander: Tipulidse from Eastern Asia, XXIII 361 

are slightly subterminal) ; tibise yellow, the tips very narrowly 
darkened ; tarsi broken. Wings (Plate 1, fig. 17) with a grayish 
tinge, the ground color vaguely brightened by paler areas, chiefly 
before origin of Rs and beyond stigma; stigmal area very faint; 
a darkened area along cord, indicated chiefly by the darker veins ; 
veins pale, except as described. Costal fringe long. Venation : 
Sci ending shortly before the origin of Rs ; basal section of Rg 
long; m-cu shortly before fork of M. 

Abdomen, including hypopygium, dark brown. Male hypopy- 
gium (Plate 3, fig. 42) with the basistyle, 6, unproduced at outer 
end. Two dististyles, the outer, od, a long, simple, blackened rod, 
curved and gently sinuous to the acute tip ; inner style, id, pro- 
foundly bifid, its outer arm a yellow fusiform structure that 
terminates in an acute spine, the surface with abundant setse; 
inner arm shorter, with marginal setae. Phallosome, p, consist- 
ing of two pairs of broadly flattened, superimposed plates. 

Habitat — Central Celebes (District Bontoe Batoe). 

Holotype, male, Latimodjong Mountains, altitude 3,800 feet, 
May 15, 1931 (Clagg). 

Gonomyia (Lipophleps) toraja is named from an aboriginal 
tribe inhabiting central Celebes. The species is readily told from 
all other regional species of the subgenus by the structure of the 
male hypopygium. This latter has a structure generally like 
that of G. (L.) kertesziana Alexander (northeastern New Gui- 
nea) , but the details are entirely diflferent. 

GONOMYIA (LIPOPHLEPS) TO ALA Bp. nov. Plate 1, fiff. 18; Plate 3. fig. 43. 

General coloration of notum dark brownish gray, the lateral 
border of the prsescutum conspicuously lighter gray; pleura 
striped longitudinally with whitish; halteres yellow; femora and 
tibiae brownish yellow; wings yellowish gray; stigmal area 
faintly indicated; Rs angulated, and, in cases, short-spurred at 
origin ; male hypopygium with three dististyles, the intermediate 
one profoundly bifid; phallosome consisting of flattened pale 
plates, without blackened points or spines. 

Male. — Length, about 2.8 to 3 millimeters ; wing, 3.4 to 3.6. 

Rostrum and palpi black. Antennse black, the pedicel more 
or less brightened; verticils (male) unusually long and delicate. 
Head brownish gray, more or less variegated by paler gray. 

Mesonotum chiefly dark brownish gray, the lateral border of 
the prsescutum conspicuously lighter gray; posterior sclerites of 
notum, including the scutellum, gray. Pleura brown, with a 
broad whitish longitudinal stripe extending from and including 

362 The Philippine Journal of Science 1935 

the fore coxae to the base of abdomen, this stripe narrowly bor- 
dered both above and beneath by still darker brown. Halteres 
yellow. Legs with the coxae brown, the fore coxae whitish, as 
described ; trochanters obscure yellow ; remainder of legs brown- 
ish yellow, the outer tarsal segments more blackened. Wings 
(Plate 1, fig. 18) with a yellowish gray tinge, the prearcular 
and costal portions a little clearer yellow; stigmal area faintly 
indicated; veins pale. Venation: Sc of moderate length, Sci 
ending a short distance before origin of Rs; Rs angulated and 
sometimes short-spurred at origin ; m-cu a short distance before 
fork of M. 

Abdomen dark brown, the pleural region paler; incisures, es- 
pecially of the outer segments, a little brightened. Male hypo- 
pygium (Plate 3, fig. 43) with the apex of basistyle, 6, unpro- 
duced. Three dististyles; outer, od, a slender, nearly straight 
rod, the apex narrowly blackened and gently curved into a spine, 
the outer margin with a series of microscopic setae ; intermediate 
style, md, profoundly bifid, the outer arm a little longer than 
the outer style, appearing as a flattened pale blade that is slightly 
arcuate; inner arm of this style much shorter, its apex a black- 
ened spine; inner style, id, elongate-oval, with numerous setae, 
none of which is evidently fasciculate. Phallosome, p, appear- 
ing as two flattened pale plates, with a slenderer median pale 
structure, the entire organ without blackened points or spines. 

Habitat. — Central Celebes (District Bontoe Batoe) . 

Holotype, male, Latimodjong Mountains, altitude 3,800 feet, 
May 15, 1931 (Clagg). Paratopotype, male. 

The specific name of this species, toala, is that of an aboriginal 
tribe. The fly is quite distinct from all other regional members 
of the subgenus in the structure of the male hypopygium, espe- 
cially the profoundly bifid intermediate dististyle and the entirely 
pale phallosome. 

LIPSOTHRIX KASHMIRICA sp. nav. Plate 1, fig. 19; Plate 3, &g. 44. 

Thorax yellow; legs yellow, the tips of femora and narrow 
bases of tibiae brownish black; wings whitish hyaline, without 
stigma; veins brown, the prearcular veins light yellow; abdo- 
minal tergites with a broad, continuous, dark brown, median 
stripe; male hypopygium with the interbase bearing two small 
acute spines on outer face before midlength. 

Male. — ^Length, 7.5 to 8 millimeters ; wing, 8 to 8.5. 

Female. — ^Length, 8.5 to 9.5 millimeters; wing, 9 to 9.5. 

56, 3 Alexander: Tipulid^ from Eastern Asia, XXIII 363 

Rostrum yellow, palpi pale. Antennae with basal segments 
pale yellow, the outer ones passing into brown ; the number of 
pale basal segments varies considerably in different specimens, 
in some cases involving several of the flagellar segments, as well 
as the scape and pedicel. Head brownish yellow. 

Thorax entirely yellow. Halteres yellow. Legs yellow, the 
femoral tips and narrower tibial bases brownish black; tips of 
tibiae narrowly darkened; outer tarsal segments infuscated. 
Wings (Plate 1, fig. 19) whitish hyaline, without stigmal or 
other markings ; veins brown, the prearcular veins light yellow. 
Macrotrichia on all longitudinal veins beyond cord, including the 
outer ends of both anal veins, the amount on the latter veins 
variable in different specimens. Venation: m-cu at or close to 
fork of M; basal section of R5 and r-m often angulated and 
weakly spurred. 

Abdominal tergites with a broad, continuous, dark brown, 
median stripe, the lateral borders of the segments yellow, more 
widely so on the outer segments; seventh to ninth segments 
(male) uniformly blackened. Male hypopygium (Plate 3, fig. 
44) with the interbase, i, long and sinuous, at base with a 
slender spinous point; on outer margin before midlength with 
two small acute spines. 

Habitat — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 9,000 feet, May 30, 
1934 (Miss Hutchinson). Allotopotype, female, altitude 8,000 
feet. May 24, 1934. Paratopotypes, 15 of both sexes, altitude 
8,000 feet. May 22 to 30, 1934. 

Lipsothrix kashmirica is most nearly allied to L. errans 
(Walker) , differing especially in the nearly hyaline wings and 
conspicuous, entire, median dark stripe on abdominal tergites. 
Edwards 2 has indicated the existence in Europe of no fewer 
than four species, of which three occur in Britain. Four other 
species are found in Japan and Formosa, but hitherto none had 
been recorded from the Asiatic mainland. 

ORMOSIA HUTCHINSONiS: sp. nov. Plate 1, fifir. 20; Plato 8, flar. 45. 

Mesonotum gray, with scarcely indicated prsescutal stripes; 
antennae with scape and pedicel pale, flagellum black; thoracic 
pleura yellow; knobs of halteres dark brown, femora yellow, the 
tips broadly blackened; wings whitish subhyaline; stigma and 

• Entomologist 63 (1930) 212. 

364 The Philippine Journal of Science ^^^^ 

very narrow seams along cord and outer end of cell 1st M2 
brown; cell 1st M2 closed; anal veins divergent; abdominal ter- 
gites brownish black, the sternites and hypopygium yellow. 

Male. — Length, 3.8 to 4.2 millimeters; wing, 4.5 to 5. 

Female. — Length, about 5 to 5.5 millimeters; wing, 5.5 to 6. 

Eostrum brownish yellow to yellow; palpi dark brown. An- 
tennae short ; scape light yellow ; pedicel brownish yellow ; flagel- 
lum black; flagellar segments oval, the outer segments a little 
more elongate. Front light yellow ; vertex light gray. 

Mesonotum gray, the praescutal stripes not or scarcely de- 
fined; tuberculate pits black, conspicuous; lateral margins of 
prsBScutum paling to yellow; lateral pretergites light yellow. 
Pleura yellow to weakly infumed, contrasting with the notum. 
Halteres dark brown, the basal portion of stem yellow. Legs 
with the coxse and trochanters light yellow; femora yellow, 
the tips broadly blackened; tibiae brown, the tips blackened; 
tarsi black. Wings (Plate 1, fig. 20) whitish subhy aline; stigma 
and very narrow seams along cord and outer end of cell 1st M2 
brown; veins brown. Venation: Cell 1st M2 closed; m-cu not 
far beyond the fork of M ; anal veins divergent. 

Abdominal tergites brownish black, sparsely pruinose ; lateral 
borders of segments very narrowly pale; sternites and hypo- 
pygium light yellow. Male hypopygium (Plate 3, fig. 45) with 
the dististyle, d, apparently simple but deeply bifid, the outer 
arm a cylindrical blackened structure that terminates in a group 
of acute teeth ; inner arm pale, broad-based, narrowed to a long 
slender point, the tip obtuse. 

Habitat. — Kashmir. 

Holotype, male, Kaj-Nag Range, altitude 8,000 feet. May 15, 
1934 {Miss Hutchinson). Allotopotype, female. Paratopo- 
types, 15 of both sexes. 

Ormosia hutchinsonse is named in honor of Miss Vivien R. 
Hutchinson, to whom I am greatly indebted for many interesting 
Tipulidse from Kashmir. The species is readily distinguished 
from all other Palaearctic species having cell 1st M2 closed by the 
coloration of the body and peculiar conformation of the dististyle 
of the male hypopygium. 

STYRINGOMYIA CELEBESENSIS sp. nov. Plate 1, figr. 21; Plate 3, fi». 46. 

General coloration yellow; mesonotum, especially the scutum, 
with specially modified setse; wings with vein R3 nearly trans- 
verse ; 2d A strongly curved to margin ; male hypopygium with 

56,3 Alexander: TipuUdse from Eastern Asia, XXIII 365 

apex of basistyle bispinous; outer arm of dististyle without 
median projection. 

Male. — Length, about 5.3 to 5.5 millimeters ; wing, 5.2 to 5.4. 

Female. — Length, about 5 millimeters ; wing, 4.5. 

Rostrum brown; palpi dark brown, the terminal segment 
paler. Antennae with the scape and pedicel dark beneath, more 
yellow above; flagellum yellow, the basal segments short and 
crowded. Head grayish brown; setae large and conspicuous. 

Pronotum whitish, with conspicuous setae. Mesonotal prae- 
scutum yellowish brown, vaguely lined with darker brown ; notal 
setae conspicuous, especially those of the scutum, much as in 
S. ensifera; mediotergite pale. Pleura testaceous-yellow. Hal- 
teres yellow. Legs with the coxae and trochanters yellow; re- 
mainder of legs yellow, the femoral and tibial rings broad but 
diffuse, brown ; fore femora with a group of strong black setae 
at tip; tarsi pale, the tips of the segments narrowly darkened. 
Wings (Plate 1, fig. 21) pale yellow, the costal border slightly 
more saturated; small dark areas on r-m; both ends of basal 
section of vein M3 but with the intermediate portion pale, m-cu ; 
marginal spots at ends of all medial, cubital, and anal veins, 
largest and most conspicuous on 2d A. Venation: R3 nearly 
transverse; cell 2d M2 narrow to broadly sessile; vein 2d A 
strongly curved into margin but without angulation. 

Abdomen obscure yellow, the caudal borders of the tergites 
narrowly darkened. Male hypopygium (Plate 3, fig. 46) with 
two broadly flattened spines on basistyle, 6, these sessile or from 
very short lobes. Outer lobe of dististyle, od, with the basal 
portion pale and slightly dilated, with abundant setae ; more than 
the outer half of arm narrowed and darkened ; no lobe at near 
midlength, as in ensifera; intermediate, md, and inner, id, arms 
of dististyle complex, especially the latter. Tenth tergite, t, 
with an elongate liguliform terminal lobe. Ninth sternite, 9s, 
broad, the usual modified outer setae distinctly subterminal in 
position, the sternite projecting strongly beyond their insertion, 
narrowed and feebly darkened at apex. 

Habitat.— Centvsil Celebes (District Bontoe Batoe). 

Holotype, male, Latimodjong Mountains, altitude, 3,800 feet. 
May 15, 1931 (Clagg). Allotopotype, female. Paratopotypes, 
1 male, 2 females. 

The other species of Styringomyia with bispinous basistyles 
(armata Edwards, claggi Alexander, ensifera Edwards) differ 

366 The Philippine Journal of Science 1935 

conspicuously from the present fly in the structure of the male 
hypopygium. The latter comes closest to ensifera, differing in 
the conformation of both inner lobes of the dististyle and the 
lack of a medium lobule on the outer arm of the style. 

STYRINGOMYIA SIBERIENSIS sp. nov. Plate 1, &g, 22; Plate 3, fig. 47. 

General coloration of mesonotum pale, variegated with darker; 
scutellum black, with a yellow central spot; wings with brown 
clouds on r-m, m-cu, outer end of cell 1st M2, and end of vein 
2d A ; vein 2d A angulated and more or less spurred near outer 
end; male hypopygium with the basistyle bearing a single lobe 
and spine; ninth sternite with two setse. 

Male. — ^Length, about 5.5 millimeters ; wing, 4.2. 

Female. — ^Length, about 5.3 millimeters; wing, 4.3. 

Rostrum and palpi brownish black. Antennae with scape and 
pedicel black; flagellum obscure yellow throughout. Head 
brownish yellow, sparsely pruinose ; occipital region with a brown 
area on either side of the median line. 

Pronotum pale medially, dark brown on sides. Mesonotal prse- 
scutum brownish gray, narrowly lined with darker brown ; scutal 
lobes brown, the centers brownish yellow; median area of scu- 
tum yellow; scutellum black with a yellow median spot; medio- 
tergite brownish black, sparsely pruinose; pleurotergite testa- 
ceous-yellow. Pleura variegated dark brown and obscure yellow. 
Halteres pale yellow throughout. Legs with the coxse and tro- 
chanters yellow; remainder of legs yellow, the femora with two 
narrow dark brown rings that are interrupted beneath; tibiae 
yellow, with two brown rings, the median one incomplete ; tarsi 
yellow, the terminal segment abruptly darkened. Wings (Plate 
1, fig. 22) yellow, with brown clouds on r-m, m-cu, outer end 
of cell 1st M2, and end of vein 2d A; veins yellow, darkened^ 
in the clouded areas. Venation: Cell 2d M2 narrow to more 
broadly sessile; vein 2d A angulated and short- to long-spurred 
at point of angulation. 

Abdominal tergites more or less bicolorous, especially in male ; 
obscure yellow in central portions, the bases and more narrow 
tips brown ; sternites uniformly pale yellow. Male hypopygium 
(Plate 3, fig. 47) with the tenth tergite, t, produced into a con- 
spicuous apical lobe; ninth sternite, 95, slender, the two setse close 
together. Basistyle, b, with the outer lobe slender, exceeding 
its terminal spine in length. Dististyle with the outer arm, od, 
provided with a series of four setae at near midlength; inner arm, 
id, with the marginal spines in three groups; from near the 

66, z Alexander: Tvpulidse from Eastern Asia, XXIII 367 

second of these groups arises a slender pale lobe that is tipped 
with a spine ; discal setae of inner arm about fourteen in number ; 
intermediate arm, md, with apex produced into a single, power- 
ful, black spine, with a comb of smaller black pegs just beneath 
it ; a further comb of pegs some distance basad of the first group. 

Habitat. — Eastern Siberia (Ussuri) . 

Holotype, male, Vinogradovka, ISB"" 50' east longitude, 43*^ 
20' north latitude, August 10, 1929 (Kiritchenko). Allotopo- 
type, female. 

The types are preserved in the museum of the Russian Academy 
of Sciences. 

As is usual in this involved genus, the present species is best 
defined by the structure of the dististyle of the male hypopygium, 
especially the armature of the intermediate and inner lobes. 
By Edwards's key to the species of Styringomyia ^ the present 
fly runs to crassicosta (Speiser), an African species. Styringo- 
myia siberiensis is allied to two other species described at this 
time (omeiensis sp. nov. and separata sp. nov.), but differs from 
both in hypopygial characters. I had earlier ^ indicated the oc- 
currence of this tropicopolitan genus in Siberia, this note being 
based upon the present record. The species is more northern 
in its distribution than any other so far made known, the most 
northerly previous record being for S. nipponensis Alexander, 
from Honshiu Island, Japan, latitude about 35° north. 

STYRINGOMYIA OMEIENSIS sp. nor. Plate 1* fiff. 23; Plate S, fig, 48. 

General coloration chiefly pale, the abdomen only slightly pat- 
terned ; wings with the usual four dark spots ; cell 2d M2 broadly 
sessile; vein 2d A angulated and spurred at outer end; male 
hypopygium with the setae of the ninth sternite slightly sep- 
arated; dististyle with inner arm terminating in a single very 
long spine, with a group of about three smaller spines at its base. 

Mate. — Length, about 5 millimeters ; wing, 4.8. 

Rostrum brownish yellow; palpi light brown. Antennae with 
the scape blackened beneath, obscure yellow above; pedicel 
brownish black; flagellum broken. Head chiefly pale. 

Mesonotal praescutum pale, lined with brownish black on an- 
terior portion; scutum chiefly pale; scutellum blackish, pale me- 
dially; mediotergite darkened. Pleura pale yellow. Halteres 
broken. Legs with the fore and middle coxae pale yellow ; femora 

•Trans. Ent Soc. London for 1914 (1914) 210-212. 
* Philip. Journ. Sci. 52 (1933) 395. 

368 The Philippine Journal of Science 1935 

pale yellow, each with two narrow dark rings that are slightly 
interrupted beneath; tibiae with the apex and a premedial ring 
dark; tarsal segments whitish, the fore tarsi with the tips of 
the individual segments narrowly darkened; terminal segment 
black. Wings (Plate 1, fig. 23) pale yellow, with the usual four 
dark spots at r-m, m-cu, outer end of cell 1st M2, and outer 
end of vein 2d A. Venation: Cell 2d M2 broadly sessile, vein 
2d A angulated and spurred at outer end. 

Abdomen chiefly pale yellow, only slightly patterned at poste- 
rior borders of the tergites. Male hypopygium (Plate 3, fig. 48) 
with the ninth sternite, 9s, relatively broad at tip, the two setae 
slightly separated at base. Armature of dististyle as shown; 
note that the apex of inner arm, id, terminates in one very long 
spine, surrounded at base by about three shorter ones; from 
this apical group the margin of the lobe slopes obliquely to base, 
with an almost continuous group of peglike spines along this 
margin ; discal spines about a dozen in number. 

Habitat. — China (Szechwan). 

Holotype, male, Mount Omei, altitude 4,500 feet, August 9, 
1929 {Franck). 

Styringomyia omeiensis is most nearly allied to S. separata 
sp. nov. and S. siberiensis sp. nov., being most readily distin- 
guished by the details of structure of the male hypopygium. As 
in the case of the species mentioned, by the use of Edwards's 
key to the species of the genus,^ the present fly runs to S. crassi- 
costa (Speiser), of Africa. 

STYRINGOMYIA SEPARATA sp. nor, Plate 1, fij. 24; Plate S, fis:. 49. 

Male, — Length, about 6 millimeters; wing, 4.5. 

Female. — Length, about 5 millimeters ; wing, 4. 

Characters much as in S. omeiensis sp. nov., to which it is 
most nearly allied, differing especially in the structure of the 
male hypopygium. 

Prsescutum almost entirely brownish yellow, the dark mark- 
ings best represented by four areas immediately before the suture 
and by the strongly darkened cephalic portion of the sclerite 
before the pseudosutural fovese, the two regions interconnected 
by narrow darkenings on the interspaces. Wings (Plate 1, fig. 
24) with vein 2d A very weakly angulated but unspurred. 
Male hypopygium (Plate 3, fig. 49) with the inner lobe, id, of 

" Trans. Ent. Soc. London for 1914 (1914) 210-212. 

56,3 Alexander: Tipulidss from Eastern Asia, XXIII 369 

the dististyle of distinctive conformation, the outer point ter- 
minating in about six powerful spines that are widely separated 
from the next group along the margin by a broad U-shaped 
notch ; discal group of spines about ten in number, all long and 
slender; at base of inner lobe on mesal face with a small pale 

Habitat, — Formosa (north) . 

Holotype, male, Rimozan, May 2, 1933 (Issiki) . 

Allotopotype, female, pinned with type. 


[a, JEdeasvts; 6, basistyle; d, dististyle; o» gonapophysis ; i, interbase; id, inner dististyle; 
md» intermediate lobe of dististyle; od, outer dististyle or outer lobe of dististyle; p, 
phallosome; «, sternite ; t, tergite.] 

Plate 1 

Fig. 1. Trichocera arisanensis sp. nov., venation. 

2. Longurio (Longurio) variceps sp. nov., venation. 

3. Tipula (Acutipula) alhoplagiata sp. nov., venation. 

4. Tipula (Schummelia) indiff evens sp. nov., venation. 

5. Tipula (Schummelia) nigrocellula sp. nov., venation. 

6. Tipula (Lunatipula) trialbosignata sp. nov., venation. 

7. Tipula gressitti sp, nov., venation. 

8. Nephrotoma atrolatera sp. nov., venation. 

9. Dolichopeza (Nesopeza) lugubrivestis sp. nov., venation. 

10. Limonia (Limonia) subhostilis sp. nov., venation. 

11. Limonia (Dicranomyia) transfuga sp. nov., venation. 

12. Dicranota (Rkapkidolabis) subsordida sp. nov., venation. 

13. Dicranota (Rhaphidolabis) uninebulosa sp. nov., venation. 

14. Dicranota (Rhaphidolabis) pallidithorax sp. nov., venation. 

15. Dicranota (Amalopina) fumicostata sp. nov., venation. 

16. Gonomyia (Gonomyia) justa sp. nov., venation. 

17. Gonomyia (Lipophleps) toraja sp. nov., venation. 

18. Gonomyia (Lipophleps) toala sp. nov., venation. 

19. Lipsothrix kashmirica sp. nov., venation. 

20. Ormosia hutchinsonse sp. nov., venation. 

21. Styringomyia celebesensis sp. nov., venation. 

22. Styringomyia siberiensis sp. nov., venation. 

23. Styringomyia omeiensis sp. nov., venation. 

24. Styringomyia separata sp. nov., venation. 

Plate 2 

Fig. 25. Tipula (Acutipula) alboplagiata sp. nov., male hypopygium, de- 

26. Tipula (Acutipula) alboplagiata sp. nov., male hypopygium, de- 


27. Tipula (Schummelia) indifferens sp. nov., male hypopygium, de- 


28. Tipula (Schummelia) indifferens sp. nov., male hypopygium, de- 


29. Tipula (Schummelia) nigrocellula sp. nov., male hypopygium, ninth 


30. Tipula (Lunatipula) trialbosignata sp. nov., male hypopygium, de- 



372 The Philippine Journal of Science 

Fig. 31. Tipula (Lunatipula) trialhosignata sp. nov., male hypopygium, 
ninth tergite. 

32. Tipula (Oreomyza) arisanensis Edwards, male hypopygium, dor- 

sal aspect. 

33. Tipula (Oreomyza) arisanensis Edwards, male hypopygium, lat- 

eral aspect. 

34. Tipula gressitti sp. nov., male hypopygium, details. 

35. Dolichopeza (Nesopeza) lugubrivestis sp. nov., male hypopygium, 


Plate 3 

Fig. 36. Limonia (Limonia) subhostilis sp. nov., male hypopygium. 

37. Dicranota (Rhaphidolabis) subsordida sp. nov., male hypopygium. 

38. Dicranota (Rhaphidolabis) uninebulosa sp. nov., male hypopy- 

gium, interbase. 

39. Dicranota (RJiaphidolabis) pallidithorax sp. nov., male hypopy- 

gium, interbase. 

40. Dicranota (Amalopina) fumicostata sp. nov., male hypopygium. 

41. Gonomyia (Gonomyia) jusia sp. nov., male hypopygium. 

42. Gonomyia (Lipophleps) toraja sp. nov., male hypopygium. 

43. Gonomyia (Lipophleps) toala sp. nov., male hypopygium. 

44. Lipsothrix kashmirica sp. nov., male hypopygium. 

45. Ormosia hutchinsonsa sp. nov., male hypopygium. 

46. Styringomyia celebesensis sp. nov., male hypopygium. 

47. Styringomyia siberiensis sp. nov., male hypopygium. 

48. Styringomyia omeiensis sp. nov., male hypopygium. 

49. Styringomyia separata sp. nov., male hypopygium. 

Alexander: Tipulid^ from Eastern Asia, XXIII.] [Philip. Journ. Scl, 56, No. 3 


Adexander: Tipulid^e from Eastern Asia, XXIII.] 

[Philip. Journ. Sci., 56, No. 3 


Alexander: Tipulid^ from Eastern Asia, XXIII.] 

[Philip. Journ. Sci , 56, No. 3 



By W. J. Baumgartner and Marcellus T. Surla 
Of the Department af Zoology, University of Kansas, Lawrence 


This study gives a detailed account of the metamorphosis of 
the spermatids in Amblycorypha oblongifolia (De Geer), a 
tettigoniid, and follows the changes taking place during the 

While cytologists have vi^orked on the spermatocyte divisions 
of many species of Tettigoniidse, only a few have published 
papers dealing with spermatid transformation. Sabatier 
(1890), Otte (1907), and Davis (1908) are among the few 
investigators who have worked on spermatid transformation. 
The accounts of Sabatier and Davis, however, are rather short 
and show an incomplete series of changes. The paper of Otte, 
as far as we know, is the only one that gives a detailed history. 


The species selected for this study is Amblycorypha oblongp- 
folia. Its spermatogenesis has not been worked out before. 
All of the material was obtained from adult individuals, collected 
in the neighborhood of the University of Kansas during July 
and August, 1932. 

Various fixing agents were tried, but Flemming's fluid and 
Benda's proved by far the best. The material was stained in 
sections 6 and 10 microns thick. Many staining combinations 
were tried, but Benda's method and Heidenhain's iron-haema- 
toxylin with a counterstain gave the most satisfactory results. 

Observations were made on both sectioned and live material, 
but mostly on the former. The "intravitam technic," developed 
by Baumgartner and Payne (1931), was used whenever ob- 
servations on live cells in the follicles were made. Smear prep- 
arations were also used for mature spermatozoa. 

290499 10 373 

374 The Philippine Journal of Science 1935 


Amblycorypha oblongifolia, one of the "false katydids," has 
paired testes. Each testis lies in the dorsal side of the abdomen, 
extending from the third to the sixth abdominal segment. In 
the adult the testis looks like a yellowish, somewhat flattened, 

Each testis is composed of many short cylindrical follicles of 
varying length. The follicles, which lie nearly parallel to each 
other, are usually larger in the middle and somewhat tapering 
at either end, depending upon the age of the individual, and 
are inclosed in a thin connective tissue containing yellowish 
pigment. Every follicle is made up of many cysts, which vary 
in size. In follicles of young individuals sectioned longitudi- 
nally, the germ cells are generally arranged in successively older 
generations, the spermatogonia lying at the blind end of the 
follicle, followed by the spermatocytes, which in turn are 
oriented in such a way that their heads point toward the blind 
end of the follicle. 


As a result of the two spermatocyte divisions, four sperma- 
tids are produced from each primary spermatocyte. Two of 
these spermatids differ from the other two only in the possession 
of a large nuclear element — ^the compact accessory chromo- 
some^ — otherwise they are structurally the same. Plate 1, figs. 
9 and 10, represents two young spermatids, one having the extra 
element and the other lacking it. Figure 14 shows a spermatid 
at a little later stage. It shows the fully formed nucleus and 
the nebenkern. Figure 16 shows the acroblast. The nucleus, 
nebenkern, and acroblast are three of the most common and 
characteristic structures of most orthopteran spermatids that 
have been studied. Before proceeding to describe in detail the 
changes undergone by these structures during the process of 
spermatid transformation, the history of their origin will be 

The nucleV;S. — ^The spermatid nucleus pictured in figs. 14 and 
15 arises from the anaphase chromosome plates of the second 
maturation division (figs. 2 and 3). During the anaphases, 
there is nothing of particular interest in the behavior of the 
chromosomes of this tettigoniid. They are distributed in the 
usual manner, one-half migrating to one pole of the spindle and 
the other half to the other pole (fig. 1), where they aggregate 

56,3 Baumgartner and Surla: Spermatid Transformation 375 

at a little later stage (fig. 2) . At this time the outlines of the 
individual chromosomes are still distinctly visible, but as the 
chromosomes fuse into a compact mass, all trace of their out- 
lines is lost to view, except that of the accessory chromosome, 
which maintains its identity. Figures 3 and 6 show two cells 
in which the accessory chromosome is protruding from each 
compact mass of chromosomes (autosomes) and extending into 
the cytoplasm. In the telophases this compact mass resulting 
from the fusion of the chromosomes begins to show signs of 
internal disintegration (fig. 8). When the two daughter cells 
separate, the process of disintegration of the chromosome mass 
is well under way, the vacuoles in each having greatly increased 
in number, and every vacuole having become larger and more 
distinct (fig. 9) . From this stage the mass of chromatin rapidly 
breaks up into patches of varying size, which in turn disintegrate 
into smaller pieces and larger granules. This breaking up of 
the mass gradually increases the size of the nucleus, which gains 
in volume at the expense of its density, while at the same time 
the substance itself becomes less chromophilic (figs. 10 to 13). 
When the nucleus reaches its maximum size (fig. 14), the nu- 
clear membrane appears and the spermatid nucleus is completed. 
A glance at figs. 9 and 14 shows the increase in size of the nucleus 
from the time of separation of the two daughter cells until the 
formation of the nuclear membrane. These illustrations also 
show the gradual loss of aflSnity for nuclear stain. 

The accessory chromosome, whenever present, remains intact 
throughout the stages of the formation of the spermatid nucleus. 
Its behavior differs markedly from that of the autosomes, in 
that it has not undergone a process of disintegration. It retains 
its compact form, and its staining reaction is similar to that of 
autosomes in metaphase. Before the formation of the nuclear 
membrane, it is usually found lying close to the periphery of 
the chromosome mass (figs. 10 to 13). After the appearance 
of this membrane, it takes a position close against its inner 
surface. It usually assumes an oval or lenticular shape. Be- 
cause of its large size and staining reaction, it is easily dis- 
tinguished from the rest of the nuclear elements, and appears 
very conspicuous in young spermatids. 

The nebenkern. — The second outstanding characteristic struc- 
ture of the spermatid is the nebenkern. In the early anaphase 
of the second maturation division, the mitochondria are found 
in the cytoplasm in the form of short rods and large granules. 

376 The Philippine Journal of Science 1935 

They are always closely crowded together midway between the 
poles, just under the cell wall (fig. 1). In the middle ana- 
phase the granular mitochondria have increased in size and 
become short rods of varying length. At this stage they are 
all in the form of short rods (fig. 2). They are closer to the 
spindle plate due to the elongation of the cytoplasm. In the next 
stage (fig. 3) many of the rods, especially those lying nearest 
the cell wall, join end to end to form long threads. These 
threads at first show the outlines of the short rods from which 
they are formed. At this stage a cross section of the cytoplasm 
below the chromosome plate (fig. 4) reveals the unequal dis- 
tribution of the mitochondria around the spindle. At a little 
later stage (figs. 5 and 6) many of the short rods, by the process 
already described, become long threads, which extend from pole 
to pole of the spindle. A polar view of the cytoplasm of a cell 
in the same stage, as indicated in fig. 6, shows almost uniform 
distribution of the mitochondria (fig. 7). It is at this time 
that the chromosome plates begin their telokinetic movements, 
and the spindle fibers begin to disappear. In late telophase (fig. 
8) the number of the long mitochondrial threads has increased 
greatly, but the number of the short rods has decreased. The 
long threads apparently become homogeneous in structure, for 
they stain uniformly. When the two daughter cells finally 
separate, the long threads, which extend from pole to pole, are 
divided transversely into approximately equal parts. The short 
rods, which lie at the interzonal region, are also divided. The 
rest of the short rods have been previously assorted to the cyto- 
plasm on the polar side of each nucleus. Hence, the separation 
of the two daughter cells also divides these meristically, with 
the result that approximately one-half of the entire mitochon- 
drial substance is assorted to each of the resulting spermatids. 

Later the mitochondria in each spermatid clump into a rather 
loose mass near the nucleus, which is always found to lie on 
that side of the spermatid where the cytoplasm is greatest and 
almost diametrically opposite the centriole (fig. 9). This mass 
rapidly shortens and condenses, first assuming a more or less 
rectangular form, which stains darker at both ends (fig. 10). 
At this stage its threadlike structure is still distinct, but disap- 
pears as it rounds out into a ball-shaped body (fig. 14). This 
spherical body, which stains homogeneously with basic stains, is 
the so-called nebenkern. 

The acroblast — The third spermatid structure is the acroblast, 
which differentiates first as a small vesicle from the fused mass 

66,3 Baumgartner and Surla: Spermatid Transformation 377 

of dictyosomes lying against the nuclear membrane. During the 
anaphases of the second spermatocyte division, the dictyosomes 
are distributed in the cytoplasm. Most of these lie outside the 
spindle plate in the early anaphase (fig. 1). At a little later 
stage (fig. 2), some have invaded the spindle plate. In the late 
anaphase (fig. 3), the greater number of them arrange them- 
selves along with the mitochondrial threads. Later, apparently 
one-half of them move toward the clear space on the equatorial 
side of each chromosome plate (figs. 5, 6, and 8). When the 
two daughter cells separate, each carries with it a certain num- 
ber of dictyosomes. The dictyosomes assorted to each cell 
migrate close to the periphery of the nucleus (fig. 9), where 
they fuse to form a comparatively large darkly staining body 
(figs. 10 to 13). At first this mass of dictyosomes is not ap- 
plied to the nucleus ; but later, as the nucleus increases in size, 
it gradually comes in contact with it (figs. 12 and 13). When 
the nuclear membrane appears, the mass of dictyosomes is found 
lying against its outer surface with a small vesicle already 
differentiating from it (fig. 14) . Soon this vesicle becomes en- 
larged and more evident (figs. 15 and 16). It is the spermatid 

The structure of individual dictyosomes may be distinguished 
in many of the stages described. They consist of a darkly stain- 
ing or chromophilic part and a nonstaining or chromophobic part. 
The chromophilic part has the shape of a crescent, without any 
visible horns extending from the ends. It only partially incloses 
the chromophobic part (figs. 1 to 4). The number of dictyo- 
somes found in the second maturation division varies from six 
to twelve, but the most probable number is the latter, as that 
is the one most frequently met with in many of the stages 

All of the stages described above are of short duration. 
Throughout the formation of the nebenkern the centriole, always 
surrounded by a hyaloplasmic area, is invariably found to be 
approximately opposite that of the forming nebenkern. 


The stage shown in fig. 14 may be taken as representing the 
completion of the spermatid as a cell and the beginning of the 
transformation of the spermatid into the mature spermatozoon. 

The nucleus. — The nucleus as previously described has reached 
its maximum size at this stage. It measures approximately twice 
the diameter of the nucleus at the time of the separation of 

378 The Philippine Journal of Science i^^s 

the two daughter cells. From this stage the nucleus slowly 
decreases in size until it reaches the stage just before it elongates 
(fig. 33). The chromatin patches continue to disintegrate into 
smaller masses, which in turn break down into small pieces and 
granules (figs. 15 to 18). Later these chromatin pieces and 
granules are distributed just beneath the nuclear membrane, 
leaving the center of the nucleus almost clear (figs. 19 to 21). 
Shortly after this, chromatin threads invade the clear space, and 
in many spermatids large chromatin patches of varying size also 
make their appearance. These large patches of chromatin are 
the products of the disintegration of the accessory chromosome. 
Figures 18 and 21 show two spermatids in which the accessory 
chromosome begins to disintegrate. The remnants of the ac- 
cessory chromosome are connected with one another by chro- 
matin threads. Many of these threads extend from the peri- 
phery of the nucleus, where the concentration of chromatin 
granules is greatest, into the central area (figs. 22 to 30) . Later 
the remnants of the accessory chromosome rapidly break down 
into small pieces, while the central vacuoles round out. Very 
often a large chromatin granule, which is a persisting remnant 
of the accessory chromosome, is found in the center of many 
of them (figs. 32 to 34). The nucleus reaches its minimum 
spherical size just before it elongates (figs. 32 and 33). Then 
it begins to elongate (figs. 34 and 35), and flattens out rapidly 
like a paddle (figs. 36 to 38) . During these processes of elonga- 
tion and flattening, the nucleus loses its symmetry, for one side 
elongates and flattens out faster than the other, which bulges 
out instead. During this stage of development the chromatin 
granules continually shift their position until they finally form 
a ring in the center of the nucleus. This ring gradually becomes 
flattened and elliptical. It is surrounded by the nonstaining or 
oxyphilic part of the nucleus, which forms the central clear space 
of the nucleus in the earlier stages of the transformation. Fig- 
ures 35 to 38 will give a much better idea of the changes in the 
nucleus than can possibly be conveyed in a detailed description. 
The elongation of the nucleus proceeds, while at the same time 
the granular structure of its basichromatin part disappears, 
with the exception of a small portion at the anterior end, which 
still remains granular (figs. 39, 43, and 44). Soon the entire 
basichromatin part becomes more and more compact and finally 
becomes homogeneous and stains very intensely in the last stages 
of the transformation. No elimination of a nuclear substance 
into the cytoplasm has been observed during the whole process. 

58,3 Baumgartner and Surla: Spermatid Transformation 379 

The final product of the transformation of the spermatid nucleus 
is the solid deeply staining fusiform sperm nucleus (fig. 47). 
Figure 48 represents a cross section of two sperm nuclei. 

The nebenkem. — ^After the mitochondria have aggregated into 
an apparently homogeneous mass (fig, 14) , a rapid process of 
vacuolization takes place. A few vacuoles appear at first just 
beneath the periphery. Each is separated from the others by a 
ring of chromophilic substance (fig. 16). Then the vacuoles 
coalesce to form a few larger vacuoles, which are separated one 
from the other by a chromophilic septum. The septa radiate 
from a darkly staining center to the outer ring (fig. 17) . These 
septa soon disappear, and a condition like that shown in fig. 18 
results. The nebenkem now consists of a chromophilic center, 
surrounded by a chromophobe part. Very often small vacuoles 
are found in the chromophilic central substance. The stage 
shown in figs. 18 and 19 persists for a long time and may be 
called the resting stage of the nebenkern. Following this it 
elongates and approaches the axial filament until it comes to 
lie parallel with it. Then its central core vacuolizes (fig. 20) 
and continues to elongate. Later it divides, one-half sheathing 
one side of the axial filament and the other half the other side 
(fig. 21). At first large vacuoles are found in its component 
parts ; but, as it elongates more and more, it breaks down into 
smaller pieces and the vacuoles disappear (figs. 25 to 27). The 
elongation seems to be associated with the lengthening of the 
axial filament, for, as the latter structure increases in length, 
the nebenkern trails it (figs. 29 and 32). In the later stages 
of the nebenkern, ''blebs" appear in it, which are pushed down 
as the spermatozoon is about to reach its final development, and 
form a large protoplasmic ball at the posterior end of the 
flagellum (fig. 44). This protoplasmic ball is finally cast off 
as the spermatozoon becomes mature. The final product in the 
metamorphosis of the nebenkern is a sheathlike structure, which 
forms the outer covering of the flagellum of the mature sper- 
matozoon (fig. 47). 

The acroblast and the acrosome. — The acroblast arises, as we 
have said, as a small vesicle differentiated from the dark mass 
of dictyosomes lying against the nuclear membrane near the 
nebenkem. This structure rapidly increases in size and assumes 
a rounded shape. It stains heavier on the side close to the cell 
wall, and is always in contact with the dictyosome mass from 
which it is differentiated (fig. 16). It then begins to migrate 

380 The Philippine Journal of Science 1935 

around the nucleus. During the process of migration it grad- 
ually increases in size (figs. 17 to 23). After its migration 
around the nucleus, it stops at its original position and remains 
stationary for a long time, and again increases in size. It ap- 
pears to stain very lightly at this stage (figs. 24 and 25) . Later 
the acroblast rapidly separates from the rest of the dictyosome 
mass, which has become diffused now, and disappears in the 
cytoplasm. Figures 26 to 28 show three stages in the separa- 
tion of the vesicular acroblast (acroblast remnant) from the 
diffused mass of dictyosomes. Figure 29 shows the remnant 
of the acroblast passing down the tail region. Meanwhile the 
rest of the dictyosome mass, which lies against the nuclear mem- 
brane, becomes diffused and loses its strong affinity for nuclear 
stains as it increases in size (fig. 25). Later it becomes dis- 
tinctly demarcated from the separating acroblast, and rounds 
out (figs. 26 and 27) . At this time vacuoles appear in it. These 
vacuoles coalesce to form larger vacuoles, which increase in size 
and produce a large vesicle in the center of which is a darkly 
staining granule (fig. 28). The vesicle and the granule found 
in it form the acrosome complex. The vesicle and the granule 
rapidly increase in size as they move toward the apex of the 
nucleus, opposite the centrioles (figs. 29 and 30). After the 
acrosome complex has reached the point opposite the centrioles, 
the cell's axis straightens. Thus, the centrioles become definitely 
located at the posterior end of the nucleus, and the acrosome 
complex at the anterior (fig. 32). The acrosome complex soon 
flattens out against the nucleus. Meanwhile the granule of the 
complex augments in size, while the clear vesicle begins to show 
distinct affinity for basic stains (fig. 33). As the nucleus 
elongates and flattens out, the acrosome complex, which now 
stains homogeneously, extends for some distance around its an- 
terior end, forming an intensely staining caplike structure (figs. 
34 to 38). Later the caplike structure (acrosome) becomes 
horseshoe-shaped (fig. 39). Then it assumes the shape shown 
in fig. 40. At a little later stage (fig. 43) a marked change oc- 
curs in it. It no longer stains uniformly, its median portion 
staining only slightly, while the part on either side plus a small 
anterior portion remain deeply stained. However, this differ- 
ence in staining reaction disappears when the acrosome passes 
into the last stages of its transformation (figs. 44 and 47), 
and the spermatid becomes a mature spermatozoon. As shown 
in figs. 44 and 47, the acrosome loses most of its strong affinity 
for stain. It is only slightly chromophilic now. Its final defin- 

56,3 Baumgartner and Surla: Spermatid Transformation 381 

itive shape is in the form of an A without the bar or like an 
inverted V. 

The cenirioles and the axial filament — ^After the separation of 
the two daughter cells in the last maturation division, the sphe- 
roidal centriole, surrounded by a hyaloplasmic area, appears on 
that side of the spermatid opposite the chondriosome mass (fig. 
9). As the chondriosome mass condenses and rounds out, the 
centriole maintains its position in relation to this structure ; that 
is, it is always found approximately opposite the chondriosome 
mass (figs. 10 to 13). After the nebenkern has been formed, 
the centriole, now in contact with the nucleus, begins to move 
toward it (fig. 14) . At this stage it is hard to tell whether or 
not the centriole has divided, because it is closely associated with 
a darkly staining mass of nuclear substance, which lies just 
beneath the nuclear membrane (figs. 16 and 17). In later 
stages, however, the centriole is seen already divided into prox- 
imal and distal parts, apparently equal in size. There is a very 
fine granule, scarcely thicker than the axial filament on which 
it is lying, a short distance from the distal centriole. This may 
be called the distal centriole derivative (figs. 18 and 19) . After 
the nebenkern has reached the axial filament, it begins to elon*- 
gate. Then the distal centriole derivative moves caudad and 
comes in contact with the elongating nebenkern at the posterior 
end (figs. 20 and 21). The centriole derivative continues to 
move posteriorly and is generally found at the end of the rapidly 
elongating nebenkern (figs. 25 and 32) up to a very late stage 
of the spermatid transformation when it is lost to view and its 
subsequent fate becomes unknown. At first the proximal and 
distal centrioles are nearly of the same size, but in later stages 
the former becomes larger than the latter (figs. 20 to 22, 27, 28, 
32, and 33) . When the nucleus begins to elongate, the centrioles 
become indistinct and are closely associated with, perhaps are, 
the same substance previously described in the earlier stages. 
This substance forms an envelope around them (figs. 34 and 
35). As the nucleus becomes paddle-shaped and the break in 
the nuclear membrane at the posterior end becomes more evident, 
each centriole appears to have divided longitudinally into two 
(figs. 36 to 38) . The position of the centrioles in relation to 
the basichromatin part of the nucleus as shown in fig. 36 lasts 
for a long time. They do not come in contact with it, but are 
separated by a space which is filled with very fine granules 
(figs. 37 to 43). In later stages the centrioles fuse, and with 

382 The Philippine Journal of Science mb 

the enveloping substance form a small compact mass (fig. 43). 
When the nucleus finally becomes solid and stains intensely, 
the fused centrioles come in contact with it at the posterior end 
to form the middle piece of the mature spermatozoon (figs. 44 
and 47). 

The spermatozoon. — ^The most mature spermatozoa found in 
our preparations made in the early part of summer have the 
following parts: 

1. A large slightly staining acrosome, which resembles an 
inverted V. It forms the most anterior end of the spermatozoon. 

2. A solid fusiform nucleus, which stains intensely with basic 
stains. There is no visible membrane surrounding it. 

3. A small compact middle piece connected anteriorly with the 
nucleus and posteriorly with the flagellum. 

4. A comparatively short flagellum surrounded by a delicate 
sheath. Its length is approximately twice that of the head. 


The metamorphosis of the spermatid into the spermatozoon 
in Amblycorypha shows remarkable changes and offers many 
points of interest. These changes resemble in a general way 
those described by Sabatier (1890) in several species of Tetti- 
goniidse, by Otte (1907) in Locusta, and by Davis (1908) in 
Steiroxys. In many details, however, it differs widely from the 
changes described by these authors. These differences, we are 
inclined to believe, are due mostly to differences of interpretation. 


The transformation of the spermatid nucleus into the sperm 
nucleus in Amblycorypha offers but one interesting point that is 
worth discussing. This is the differentiation of the nuclear sub- 
stance into two distinct substances, one of which shows a strong 
affinity for basic stains, the other little or none at all. The for- 
mer substance is called basichromatin and the latter oxy- 
chromatin. This differentiation, which occurs before the 
condensation of the nucleus to form the compact deeply staining 
sperm nucleus, seems to be of widespread occurrence in Tetti- 
goniidae. It has been partly described by Otte (1907) and by 
Davis (1908). The same differentiation has been described by 
Montgomery (1911) in Euschistus, by Bowen (1920) in some 
species of Hemiptera, and by others in other forms. 

50,8 Baumgartner and Surla: Spermatid Transformation 383 

Now and again a remarkable phenomenon in the transforma- 
tion of the spermatid nucleus has been reported by several in- 
vestigators of insect spermatogenesis. This phenomenon is the 
extrusion into the cytoplasm of a lightly staining substance. 
Montgomery (1911) and Bowen (1922) have observed and 
described it. This interesting phenomenon, while perhaps of 
widespread occurrence in Hemiptera, does not seem to occur in 
Orthoptera. At least, it has not been reported in the spermio- 
genesis of many orthopterans thus far examined; and as far 
as our observations on Amblycorypha are concerned, there is 
nothing to indicate that such a phenomenon occurs in the meta- 
morphosis of the spermatid in this insect. 


The extensive literature on the origin and behavior of this 
structure has been so thoroughly reviewed by recent investiga- 
tors, notably Bowen (1922), that it will not be necessary to 
present a detailed discussion here. 

In general there are two opinions held concerning the origin 
of this characteristic structure in the spermatids of insects. 
One view maintains that the nebenkern arises from the fusion 
of the chondriosomes assorted to each spermatid ; the other, that 
it originates from the direct metamorphosis of the interzonal 
filaments. The former view is supported by the works of such 
prominent investigators as Otte (1907), on Locusta; Mont- 
gomery (1911), on Euschistus; Payne (1917), on Gryllotalpa; 
and recently by the works of Bowen (1922), on several species 
of Hemiptera-Heteroptera ; PoUister (1930), on Gerris; Johnson 
(1931), on Oecanthus; and M. A. Payne (1933), on several spe- 
cies of grasshoppers. The latter view is supported by the re- 
searches of Butschli (1871), on several species of insects; La 
Vallette St. George (1886), on Blatta; Paulmier (1899), on Ana^ 
sa; and Munson (1906), on Papilio. 

Our observations are in line with the view held by the first 
group of investigators. The nebenkern in this tettigoniid is 
derived entirely from the fusion and subsequent condensation 
of the chondriosomes. 

Sabatier (1890) seems to be one of the first to observe the 
nebenkern in the tettigoniid spermatid. He believes this struc- 
ture arises from the protoplasm (cytoplasm) and calls it "vesi- 
cule protoplasmique." What finally becomes of this vesicle, he 

384 The Philippine Journal of Science 1935 

does not say. Since in his brief paper he neither gives figures 
nor describes the changes following the formation of this struc- 
ture, it is impossible to compare the behavior of his "vesicule 
protoplasmique" with the behavior of the nebenkern in Ambly- 

Otte (1907) describes in detail the origin and behavior of the 
nebenkern in Locvsta. According to his account, the greater 
portion of the mitochondria assorted to each spermatid fuses into 
a loose mass near the side of the interzonal filaments. This 
mass of mitochondria becomes condensed and forms the com- 
pact nebenkern, which stains darker than the cytoplasm. Later 
the nebenkern differentiates into two substances; namely, an 
inner chromophilic core and an outer chromophobic part. From 
the chromophilic center septa radiate and extend to the periphery 
of the chromophobic part. Then the nebenkern becomes homo- 
geneous and finally elongates to form a sheath around the axial 

Davis (1908) fails to identify the nebenkern in the spermatids 
of Steiroxys, nor does he trace the origin of a structure that 
corresponds to it. He calls "Nebenkern'* for convenience and 
"without implying anj^hing in regard to its homologies," a 
darkly staining structure which lies near the nucleus (see his 
fig. 203.) If his "Nebenkern" corresponds to the acroblast of 
recent investigators, then the less deeply staining structure ap- 
plied to the nuclear membrane and lying near the "Nebenkern" 
is probably the true nebenkern. However, if we were to inter- 
pret his figure 203, with reference to the differential staining of 
the two structures and their relative position in the cytoplasm, 
we would say the nebenkern is the same one which he calls 
"Nebenkern" (acroblast) , and the less deeply staining body close- 
ly applied to the nuclear membrane is the acroblast. 

Our observations on the origin and behavior of the nebenkern 
in Aniblycorypha agree in general and in many particulars with 
those of Otte in Locusta. However, he has not mentioned a few 
of the early stages in the differentiation of the nebenkern into 
two substances. In these particulars our account resembles 
closely the description given by Bowen (1922), of Murgantia, 
and partly that given by Johnson (1931), of Oecanthus. The 
differentiation of the nebenkern in some hemipters, as given by 
Bowen, with the exception of the initial stage — namely, the ap- 
pearance of many vacuoles in the periphery of the nebenkern — 

56,3 Baumgartner and Surla: Spermatid Transformation 385 

is so strikingly like that of Amblycorypha that it could be 
applied almost directly. Bowen states that — 

. . . once fully completed, the nebenkern begins to show the first signs of 
differentiation into two substances. This first makes itself evident by the 
appearance of a large number of vacuoles in the periphery of the mito- 
chondrial mass. These are at first very indistinct and difficult to demon- 
strate, but they presently become clearer and form a complete layer invest- 
ing the periphery of the nebenkern. They seem to stain little or not at 
all in contradistinction to the central mass, which is more darkly colored. 
This may be due to the fact that the process of differentiation is going 
on from without inward, or it may be due to the optical superposition of 
several layers of vacuoles. The outermost layer clears up very rapidly, 
the separate vacuoles fusing together to form large clear spaces, separated 
from each other by septa which pass outward from the central mass to 
the exterior of the nebenkern which is marked by a very definite mem- 
brane. The whole thing reminds one of the figure given by Meves (1900) 
of Pygaera. The clearing up of the outer layer of vacuoles now proceeds 
rapidly, the substance of the vacuole walls apparently withdrawn into the 
underlying region of the nebenkern, and eventually the whole peripheral 
zone appears as a clear non-staining envelope, enclosing a central chromo- 
philic core. 

While the processes of the differentiation of the nebenkern 
into a chromophilic central substance and a chromophobic outer 
substance in Amblycorypha resemble more closely those in He- 
miptera than those in Locusta as described by Otte, the subse- 
quent changes in Amblycorypha are entirely different from 
those observed by Bowen and Johnson, but are similar to those 
described by Otte. As a whole the changes undergone by the 
nebenkern in Amblycorypha, from the time of its formation to 
its final fate, resemble more closely those of Locusta than those 
described either by Bowen in Hemiptera or by Johnson in Oecavr- 
thus. This resemblance is not surprising at all — in fact, it is 
what one would expect to find in such closely related genera as 
Locusta and Amblycorypha. It appears very probable that, if 
a more-detailed study of the nebenkern in Locusta, in the forms 
examined by Sabatier and in Steiroxys studied by Davis, were 
made, a condition similar to that in Amblycorypha would be 


No structure in the spermatids of insects has received more 
consideration in the past than that which forms the apex of 
the sperm head, commonly called the acrosome or perforatorium. 
This structure, which varies greatly in size and shape in diverse 

386 The Philippine Journal of Science i»3b 

groups of animals, has been traced to many different origins. 
In insects alone it has been described as originating from the 
nucleus, spindle remnants, nebenkern, Golgi bodies, idiosome, 
acroblast, sphere, from a combination of mitochondria and cen- 
tral spindle fibers, from a centrosome, and in some cases it 
has been derived from an obscure structure in the cytoplasm, 
the nature of which is not known. 

Butschli (1871) appears to be the first to describe the 
formation of the acrosome in the tettigoniid spermatids. He 
derives this structure from a granulelike vesicle in the cyto- 
plasm. This vacuole comes in contact with the nucleus and 
moves forward to its anterior portion to form the definitive 
acrosome. His observations are in line with those of many 
recent investigators. He fails, however, to make out the origin 
of this granulelike vesicle and to notice the casting off of a 
substance which corresponds to the so-called acroblast remnant. 

Sabatier (1890) in his studies of the spermatogenesis of some 
locustids traces the origin of the acrosome to the **vesicules 
nucleaires," which arise from the direct metamorphosis of the 
nucleus. A fusion occurs among these small nuclear vesicles, 
and as a result a small number of large vesicles is produced. 
These large vesicles increase in size, lose gradually their affinity 
for stain, and diminish by a further fusion to three; namely, 
"un mediane, petite et saillante, et deux laterales qui s'allongent 
et finissent par former les deux branches ou crochets de I'ancre 
que constituent la coiffe cephalique." 

It seems probable that Sabatier misinterprets the complicated 
changes occurring during the transformation of the spermatid 
nucleus in the locustid forms he has examined. His observa- 
tions on the origin of the acrosome are entirely different from 
our observations on Amblycorypha and from those of others 
who have later made careful studies on the origin of this struc- 
ture in closely related insects. 

Otte (1907) gives a complete, as well as a most satisfactory, 
account of the formation of the acrosome in Locusta. Our ac- 
count of the origin of the acrosome in Amblycorypha, with the 
exception of two important points, is in accord with his on 
Locusta. In his observations on the origin of the idiosome, 
the structure in the spermatid from which the acrosome origi- 
nates, he describes it as arising from the direct metamorphosis 
of the bundle of central spindle fibers (interzonal filaments) and 
a portion of the mitochondria. As far as our observations are 

66,3 Baumgartner and Surla: Spermatid Transformation 387 

concerned, we are quite sure that the central spindle fibers and 
a portion of mitochondria take no part in the formation of the 
acroblast in Amblycorypha. Here the central spindle fibers dis- 
appear after the telokinetic movements of the chromosome plates 
have taken place ; while the acroblast makes its appearance only 
after it has been differentiated from the fused mass of dictyo- 
somes lying against the nuclear membrane, long after the dis- 
appearance of the spindle fibers and the aggregation of the 
mitochondria into a compact spherical body. Recent investiga- 
tions on this subject show conclusively that the idiosome or 
acroblast does not arise from the spindle fibers and mitochon- 
dria but from other structures in the cell, such as Golgi bodies 
and dictyosomes. 

Again Otte describes the casting off of the mitochondrial sub- 
stance from the idiosome after its rotation around the nucleus. 
In Amblycorypha the substance that is cast off from the forming 
acrosome is the rounded vesicle (acroblast remnant), which 
corresponds to his idiosome. In short the idiosome itself forms 
the acrosome in Locusta, while in Amblycorypha the acrosome 
is formed by the acrosome complex synthesized in close connec- 
tion with the acroblast. 

Davis (1908) describes the acrosome in Steiroxys as arising 
from a rounded structure which he calls, for convenience, "Ne- 
benkern." The nature of this body he does not know. He 
traces the transformation of this structure into the anchor- 
shaped acrosome. He has not observed, however, the differen- 
tiation of the acrosome complex in close association with the 
"Nebenkern,'' and the casting off of a substance from it. He 
also overlooks some of the later stages of the acrosome. 

Recently the origin of the acroblast and its role in the forma- 
tion of the acrosome have been more fully studied in other 
insects that are not closely related to Amblycorypha. A brief 
review of the results of a few of these investigations will be 
given here for comparison with our findings. 

Payne (1916), in his study of the germ cells of Gryllotalpa, 
like Davis, is not able to ascertain the origin of the acroblast. 
He observes two structures in the cytoplasm of older spermatids, 
one of which is elongated and curved and is in contact with the 
nuclear membrane, the other small and spherical. The elongated 
body gives rise to the acrosome, while the spherical one degen- 
erates and disappears during the transformation. Concerning 
the origin of these structures, Payne says, "it would seem that 

388 The Philippine Journal of Science i98& 

they rise de novo in the cjrtoplasm as they are not present in 
the young spermatids." 

Bowen (1920 and 1922), in his study of the spermatogenesis 
of many hemipters describes the fused type of acroblast as 
originating from the fusion of the dictyosomes assorted to each 
spermatid. From the acroblast the acrosome complex is differ- 
entiated, after which the acroblast remnant is cast off from 
the acrosome complex. This complex, which consists of the 
acrosomic vesicle and the deeply staining granule found in it, 
gives rise to the definitive acrosome. This differentiation of 
the acrosome from the acroblast seems to be of very wide ap- 
plication in insects. It has been found in many of the insects 
thus far examined in detail. 

Voinov (1925) describes the formation of the acrosome in 
Gryllotalpa vulgaris. It arises, according to him, from two 
structures ; namely, the acroblast and the **appareil spherulaire.'' 
The acroblast arises in close association with the Golgi bodies, 
while the "appareil spherulaire" arises by the fusion of spherical 
bodies in the spermatid. This process of acrosome formation 
is slightly different from the condition observed by Bowen. 
However, both involve the acroblast as participating in the 
formation of the acrosome. 

F. Payne (1927) in another paper finds that the acrosome 
originates from a large idiosome. According to his account, 
several small spheres, which he calls proidiosomal spheres, fuse 
to form a large idiosome, around which the Golgi bodies collect 
or aggregate. This complex corresponds to the acroblast of 
other investigators. However, in Gelastocoris the large idiosome 
alone is responsible for th^ formation of the acrosome, the Golgi 
bodies being cast off. Thus, while the acrosome is the product 
of the acroblast in the forms studied by Bowen, in Gelastocoris 
it is the product of the large idiosome. In both cases a sub- 
stance is cast off from the structure that finally forms the 
definitive acrosome. 

In Gerris, PoUister (1930) traces the origin of the acrosome 
from the acroblast. He observes that the osmiophilic Golgi 
materials aggregate in the young spermatid and form "an 
irregularly folded, sac-like structure," the acroblast. After the 
formation of the acrosome, the acroblast remnant separates 
from it and is cast off into the cytoplasm. 

Johnson (1931) finds in Oecanthus a condition nearly like that 
described by Bowen. The acroblast is produced by the fusion 
in the spermatids of a number of dictyosomes, from which the 

56,3 Baumgartner and Surla: Spermatid Transformation 389 

acrosome is apparently differentiated. Later the acroblast rem- 
nant is cast off into the cytoplasm and passes down the tail. 

Our study of the results of the above investigators and many 
others reveals that in many cases the acrosome in insects is 
formed in close association with the acroblast or a structure 
similar to it. 


The behavior of the centrioles during the transformation of 
the spermatid is so different in various groups of animals that 
only observations on closely related genera will be considered in 
this brief discussion. Those desiring a lengthy discussion on 
this particular subject are referred to the works of Bowen 
(1922), Baumgartner (1929), and Johnson (1931). 

Early workers on the spermiogenesis of the tettigoniids have 
failed to demonstrate the presence of centrioles in the spermatids. 
Perhaps this is not because centrioles are not present in the 
objects examined, but because they are overlooked or considered 
insignificant or because the technic used at the time is not ade- 
quate to bring out such minute structures. 

Otte (1907) seems to be the first to demonstrate the cen- 
trioles in the locustid spermatids. He traces the behavior of 
these bodies throughout the transformation of the spermatid. 
The centrioles, according to his account, are so small that they 
are not easily recognized. He finds two small granules near 
the periphery of the cell, which he assumes to be central bodies. 
Later these are oriented in proximal and distal positions, the 
larger proximal centriole lies at the nuclear membrane and the 
distal one at the cell wall. Between them is the intracellular 
axial filament and extending from the distal centriole is the 
extracellular axial filament. The proximal centriole is usually 
so closely associated with the accessory chromosome or with 
chromatin patches that it cannot be recognized. Later this as- 
sociation becomes so close that it gives one the impression, "als 
kame der Achsenfaden von dem akzessorischen Chromosom." 
When the accessory chromosome breaks down, the proximal cen- 
triole emerges and then divides laterally into two. Each of the 
later divides, giving rise to four centrioles. The two inner cen- 
trioles move up into the nucleus, and each again divides into 
two. These four centrioles resulting from the division of the 
two, migrate farther to the anterior pole of the nucleus and 
form an "Innenkorper" in the sperm nucleus. Meanwhile the 

290499 11 

390 The Philippine Journal of Science 1935 

two centrioles left at the nuclear membrane elongate and take 
part in the formation of the middle piece. The distal centriole 
continues to move caudad until it disappears from view in the 
last few stages of the transformation. 

The behavior of the centrioles in Amblycorypha differs greatly 
from that described by Otte for Locusta. In both genera the 
centrioles are similar. In both chromatin material is in close 
association with the centrioles in the early stages of the trans- 
formation ; in both the axial filament originates from one of the 
centrioles; the caudal migration and final disappearance of a 
granulelike structure in close association with the axial filament 
take place in both. They are also similar in the appearance of 
four centrioles just as the spermatid nucleus begins to elongate 
and in the participation of the centrioles in the formation of 
the middle piece. Otte's recognition of the granulelike structure, 
which migrates caudad and finally disappears, is significant. 
This structure he interprets to be the entire distal centriole, 
while we believe it is only a derivative of the distal centriole. 
Recently Johnson (1931) finds in Locvsta viridissima, the spe- 
cies studied by Otte, that it is not the entire distal centriole 
which migrates caudad as Otte thought, but only a derivative of 
this. This being the case, the four centrioles, which Otte 
believed to have arisen from the proximal centriole, must have 
come from the division of the original centriole into proximal 
and distal ones, each of which in turn divides into two, giving 
rise to four, as in Amblycorypha. His description of the migra- 
tion of two centrioles up into the nucleus, their subsequent divi- 
sion into four, and their final fate to form the "Innenkorper* ' in 
the sperm nucleus, we cannot correlate with our observations. 
These four centrioles in the nucleus of the older spermatids in 
Locusta look more like some basichromatin granules in the 
center of many clear areas in the nucleus of the spermatid in 

The behavior of the centrioles in Steiroxys, as observed by 
Davis (1908), is almost identical with that of Amblycorypha. 
However, Davis seems to have overlooked the presence of cen- 
trioles in the young spermatids. It is only when the nucleus 
begins to elongate that he notes the presence of these structures. 
Hence he does not determine the origin of the axial filament 
nor does he notice the presence of a granulelike structure lying 
on it or in close association with it. It seems probable that, 
if a further study of the centrioles in Steiroxys and Locusta were 

56,3 Baumgartner and Surla: Spermatid Transformation 391 

made, a condition much more like that in Amblycorypha might 
be found. 


1. The mass of chromosomes in each daughter cell after the 
last maturation division breaks up into patches, which gradually 
become diffused through the nucleus. Later a differentiation 
of the nuclear material into a basichromatin substance and an 
oxyphilic substance takes place. From the basichromatin sub- 
stance is formed the solid, deeply staining, fusiform sperm 

2. The mitochondrial substance assorted to each spermatid 
clumps into a mass. This mass rapidly condenses and finally 
rounds out to form the spherical nebenkern. The nebenkern, 
after undergoing many changes, differentiates into an inner 
chromophilic core and an outer chromophobic part, surrounded 
by a well-marked membrane. Later it comes to lie on or below 
the axial filament, elongates with this structure, and finally forms 
a covering around the flagellum. 

3. The dictysomes in each spermatid fuse. The mass result- 
ing from this fusion comes to lie against the nuclear membrane 
near the nebenkern. From it a vesicle is differentiated which 
increases rapidly in size to form the acroblast. The acroblast 
migrates around the nucleus, and, after the formation of the 
acrosome complex in close association with it, it is finally cast 
off into the cytoplasm. The acrosome complex, which consists 
of a vesicle containing a deeply staining granule, becomes the 
definitive acrosome, having the shape of an inverted V. 

4. The spherical centriole divides into proximal and distal 
parts. From the distal centriole a derivative arises, which 
migrates caudad and finally disappears from view in the last 
stages of the transformation. Later the proximal and distal 
centrioles divide. The four centrioles resulting from this divi- 
sion participate in the formation of the middle piece. 

5. The axial filament apparently originates from the central 
bodies, but it has not been determined whether or not it orig- 
inates from the distal centriole. From it the comparatively 
short flagellum of the mature spermatozo5n is formed. 

6. The spermatozoon consists of a large inverted V-shaped 
acrosome, which stains lightly or not at all; a solid, deeply 
staining, fusiform nucleus, to the posterior end of which is 
attached the scarcely distinguishable compact middle piece ; and 

392 The Philippine Journal of Science 

a short flagellum, which measures approximately twice that of 
the head* 


Baumgartner, W. J. Spermatid Transformations in Gryllus assimilis, with 

Special Reference to the Nebenkern. Kans. Univ. Sci. Bull. 1 (1902). 
Baumgartner, W. J. Die Spermatogenese bei einer Grille, Nemobius fas- 

ciatus. Zeitschr. Mikr. Anat. 9 (1929). 
Baumgartner, W. J., and M. A. Payne. Intravitam technic used in stu- 
dies on the living cells of grasshoppers. Journ. Exp. Zool. 59 (1931). 
BowBN, R. H. Studies on insect spermatogenesis I. The history of the 

cytoplasmic components of the sperm in Hemiptera. Biol. Bull. 39 

BOWEN, R. H. Studies on insect spermatogenesis II. Journ. Morph. 

Z7 (1922). 
BoWEN, R. H. Studies on insect spermatogenesis III. On the struc- 
ture of the Nebenkern. Biol. Bull. 42 (1922). 
BUTSCHLI, 0. Nahere Mittheilungen iiber die Entwicklung und den 

Bau der Samenfaden der Insecten. Zeitschr. f. wiss Zool. 21 (1871). 
Davis, H. S. Spermatogenesis in Acridiidae and Locustidae. Bull. Mus. 

Comp. Zool. 53 (1908). 
Johnson, H. H. Centrioles and other cytoplasmic components of the male 

germ cells of the Gryllidae. Zeitschr. f. wiss. Zool. 140 (1931). 
La Vallette, St. George. Spermatogische Beitrage. Zweite Mittheilung. 

Arch. f. Mikr. Anat. 27 (1886). 
MOHR, 0. Samenreifung und Samenbilding von Locusta viridissima. 

Arch. Biol. 29 (1914). 
Montgomery, T. H, The spermatogenesis of an hemipteron, Euschistus. 

Journ. Morph. 22 (1911). 
Ottb, M. Samenreifung und Samenbildung bei Locusta viridissima. Zool. 

Jahrb. Abt. Anat. 24 (1907). 
Paulmier, F. C. The spermatogenesis of Anasa tristis. Journ. Morph. 

13 (1899). 
Payne, F. A study of the germ cells of Gryllotalpa borealis and Gryllo- 

talpa vulgaris. Journ. Morph. 28 (1916). 
Payne, F. Some cjrtoplasmic structures in the male germ cells of Gelas- 

tocoris oculatus. Journ. Morph. 43 (1927). 
Payne, M. A. Are mitochondria dynamic cytoplasmic structures? Trana. 

Kans. Acad. 36 (1933). 
Pollister, a. W. Cytoplasmic phenomena in the spermatogenesis of Ger- 

ris. Journ. Morph. 49 (1930). 
Sabatier, a. De la spermatogenese chez les Locustides. Compt. Rend. 

des Seances de L'Acad. de Sci. Ill (1890). 
VoiNOV, D. M. Les elements sexuels de Gryllotalpa vulgaris. Arch, Zool. 

Exper. et Gen. 63 (1925). 
Wilson, E. B. The Cell in Development and Heredity. Macmillan Co., 

New York (1928). 


[All fisures were drawn with a camera lucida at an initial magnification of 2,700, and the 
drawings retouched freehand.] 

Amblycorypha oblongifolia (Db Geer) 
Plate 1 

Figs. 1 to 3. Anaphases af the second maturation division. Mitochon- 
dria near the cell wall and dictyosomes scattered along the spin- 
dle plate. 
Fig. 4. Cross section of cytoplasm at the stage shown in fig. 3, showing the 
distribution of mitochondria. 

5. Late anaphase; showing short and long mitochondrial threads, also 

the small dictyosomes at both ends of the disappearing spindle 

6. Telophase; the number of long mitochondrial threads increases. 

7. Cross section of cytoplasm at the stage shown in fig. 6, showing 

the distribution of mitochondria. 

8. Telophase. 

Figs. 9 to 14. Successive stages in the condensation of mitochondrial sub- 
stance to form the nebenkern. Note also the dictyosomes. 

Fig. 15. Early spermatid; showing nebenkern, axial filament, and small 
vesicle differentiating from the dictyosome mass. 

Figs. 16 and 17. Two spermatids with fully formed acroblast; nebenkern 
differentiating into two substances. 
18 and 19. Spermatids at a little later stage; centriole divides into 
proximal and distal parts. Note centriole derivative a short dis- 
tance from the distal centriole. 
20 to 24. Later spermatids; nebenkern starts elongating; centriole 
derivative at its posterior end; acroblast moving around the 
nucleus and increasing in size. 

Plate 2 

Fig. 25. Late spermatid; dictyosome mass becomes diffused from which 

acroblast begins to separate. 
Figs. 26 to 29. Stages in the formation of acrosome complex. 

30 to 33. Note increase in size of acrosome complex. 
Fig. 34. Elongation of nucleus; flattening of acrosome complex. 
Figs. 35 and 36. Later stages. Note caplike acrosome and paddle-shaped 
nucleus and division of centrioles. 
37 to 39. Slightly later stages. 
Fig. 40. Note nucleus and acrosome; basichromatin of nucleus becoming 

Figs. 41 and 42. Cross sections of flagellum and nucleus, respectively, at 
the stage shown in fig. 40. 


394 The Philipj^ne Journal af Science 

Fio. 43. Very late stage in the transformation. Note change in staining 
reaction of acrosome. 

44. Nearly mature spermatozoon ; acrosome stains lightly and uniform- 

ly; nucleus very compact except at anterior portion. Note pro- 
toplasmic ball at posterior end of flagellum and also undulating 
membrane on each side of the nucleus. 

45. Side iriew of spermatozoon at the stage shown in fig, 44. 

46. Cross section of solid nucleus. 

47. Mature spermatozoon. 

Pigs. 48 and 49. Cross sections of nucleus and flagellum, respectively, of 
mature spermatozoon. 

Baumgartner and Surla: Spermatid Transformation.] [Philip. Journ. Scl, 56, No. 3 


Baumgartner and Surla: Spermatid Transformation.] [Philip. Journ. Sci., 56, No, 



By Karl E. Scheidl 
Of Vienna, Austria 


Female. — Nearly black, 6.20 mm long, 4 times as long as wide. 

Front flat, epistomal margin and the median line up to the 
center shining, polished, and with a lew small punctures; an- 
terior half densely punctured, subopaque and areolate above. 

Pronotum shining, 1.29 times as long as wide, widest behind 
the lateral emarginations ; the latter shallow; median sulcus long 
and fine, surrounded by cordiform patch of densely placed and 
uniform-sized punctures, the patch longer than wide; surface 
with small, sparingly placed punctures. 

Elytra but little wider than the pronotum (16 : 15.5) and 
2.31 times as long, sides parallel, rather narrowly rounded be- 
hind ; striate-punctate, with the exception of the eighth all strise 
impressed, punctures of the first and ninth striae confluent on 
the greater part of their length, striae two to seven with rather 
large and deep punctures, the eighth striae with the punctures 
small, shallow and remotely placed, interstices subconvex and 
impunctate, base of the third widened and with a long row of 
densely placed transverse rugae ; declivity rugose and hairy, con- 
vex above, perpendicular below. Abdomen normal. 

Types in the possession of Mr. F. C. Hadden and in my own 
collection. The female is here described for the first time. 

Locality.— M.ount Maquiling, Laguna Province, Philippine Is- 
lands (Bautista). Under bark of malugai, Pometia pinnata 


. Ma.!!e.— Nearly black, 6,30 mm long, 3.04 times as long as wide. 
A very distinct species of the group PMijpi sulcatl 

Front opaque, depressed medially, densely areolate; antennal 
scape longer than wide. 

Pronotum quadrate, shining, widest behind the moderately 
deep lateral emarginations, coarsely moderately densely punc- 


396 The Philippine Journal of Science i^ss 

tured; median sulcus fine and long, in front of the sulcus sur- 
face impunctate. 

Elytra as wide and 1.95 times as long as the pronotum, sides 
parallel, broadly rounded behind, striate-punctate, striae very 
narrow and very deeply impressed on the anterior two-thirds, 
becoming shallower to obsolescent towards the apex, interstices 
convex near the base, gradually flattening out caudad; in front 
of the first four interstices a narrow transverse area covered 
with transverse rugae, the interstices shining and sparsely punc- 
tured on the basal fourth of their entire length, opaque and 
rather densely covered with irregularly placed, deep, shining 
punctures behind, the punctures becoming more numerous and 
shallower towards the apex, with rows of reddish hairs; cy- 
lindrical, declivity convex and flattened on an equilateral trian- 
gular areas as in some species of Xyleborus, near the upper 
corner with a granule in the prolongation of the second 
interstice; a few more but smaller granules on the flattened 
portion below. Abdomen normal. 

Type in my collection. 

Locality. — Philippine Islands. Collected from anang, Diospy- 
ros ahemii Merrill. 


Female. — Dark reddish brown, the anterior half of the elytra 
paler, 6 mm long, 2.95 times as long as wide. This species 
shows characters of the pronotum that have not been known 
in other species of the family. Preliminarily I place it in the 
Platypi sulcati. 

Front plano-concave, depressed in the center, subopaque, sub- 
impunctate on the anterior third, areolate above, lateral margin 
towards the articulation of the antennae raised and longitudinally 
wrinkled. Antennal scape longer than wide. 

Pronotum quadrate, lateral emarginations weakly developed, 
surface shining and with scattered fine punctures; median sul- 
cus long, extending nearly to the middle of the pronotum, with 
a cordiform patch of rather large, deep punctures around it; the 
patch much longer than wide, transverse in front, anterior to 
this patch with a group of nine to twelve large pores on each 
side of the median line, the outer line of these pores is semi- 
circular on one side, nearly quadrate on the other. 

Elytra little wider than the pronotum (21 : 19.5) and 2.05 
times as long, the base feebly carinate, sides subparallel, broadly 
rounded behind; surface shining on the anterior two-thirds, 

8«' 3 Schedl: Fauna Philippinensis, III 397 

subshining behind; sulcate, sulci narrow, irregularly uniseriately 
punctured, disappearing towards the declivity; interstices con- 
vex on the shining portion, flat behind, with fine, widely scat- 
tered punctures; the ninth interspace densely, roughly punc- 
tured, base of interstices two to five transversely rugose, more 
strongly on the third; declivity commencing behind the anterior 
two-thirds, feebly convex at first, perpendicular behind, the per- 
pendicular face triangular, the entire declivity covered with ex- 
tremely crowded small shining punctures on an opaque surface 
and with numerous short yellow hairs. Abdomen normal. 

Type in my collection. 

Locality. — ^Philippine Islands. From lumbayao, Tarrietia /a- 
vanica Blume. 


Male. — Dark reddish brown, declivity black, 4.56 mm long, 
3.13 times as long as wide. Belonging to the group Crossotarsi 
subdepressi this species is easily recognized by the noncarinate 
interstices on the upper convexity of the declivity and the char- 
acters of the last abdominal sternite. The latter is opaque and 
has a rather deep emargination on the caudal margin, thus form- 
ing a lateral bluntly rounded edge on each side. 

Front narrowly depressed above the epistomal margin, con- 
vex above, entirely opaque, with a small shining puncture in 
the center and a few scattered punctures towards the vertex. 

Pronotum brightly shining, slightly longer than wide 
15 : 14), widest behind the rather shallow lateral emarginations ; 
median sulcus weakly developed, with scattered fine punctures; 
a few coarser ones at both sides of the median line in front of 
the sulcus, densely but finely punctured on a narrow strip along 
the basal margin. 

Elytra slightly wider than the pronotum (15 : 14) and 1.80 
times as long, widest a short distance behind the middle, sides 
subparallel, broadly rounded and with a semicircular emargina- 
tion behind; base carinate, surface shining, striate-punctate, 
strise impressed, less so laterad, strial punctures small and con- 
fluent on the first five striae, punctures little more remotely 
placed at the sides, interspaces flat, with scattered fine punc- 
tures on the disc, punctures becoming very densely placed to- 
wards the base and the sides ; declivity commencing in the pos- 
terior third, uniformly strongly convex, with a lunate nearly 
perpendicular depression below; the shining interstices, which 
are of equal width, cease suddenly but without any armature 

398 The Philippine Journal of Science 1935 

at the upper limit of the declivity ; the latter opaque with rows 
of yellow hairs, the upper limit of the lunate depression devel- 
oped as a feebly raised line; the lateral processes with one or 
two small spinules on the outer margin. 

Female. — Reddish brown, pronotum and basal part of the 
eljrtra paler, 4.75 mm long, 3.37 times as long as wide. 

Front flat, with a hump-shaped elevation in the center; sur- 
face opaque, with few scattered punctures in front and shallow 
strigose punctures behind; angle separating the front from the 
vertex acute. 

Pronotum longer than wide (15.5 :13.5), lateral emargina- 
tions shallow, surface shining, minutely reticulate, and with 
widely scattered punctures; median sulcus fine and long, sur- 
rounded by an oval patch of densely placed punctures, with a 
depression on each side of the median line in front of the sulcus. 

Elytra wider than pronotum (14.5 : 13.5) and twice as long, 
sides parallel, with a lunate emargination behind; sculpture as 
in the male but the punctuation not so deep; the lateral proc- 
esses shorter and simple. Abdomen normal. 

Types in the possession of Mr. F. C. Hadden and in my col- 

Localities. — Mount Maquiling, Laguna Province, and Que- 
zon Park, Tayabas Province, Luzon, Philippine Islands; F. C. 
Hadden, collector. From Ficus sp., and hagimit, Ficits minnor- 
hassae (Teysm. and De Vr.) Miq. 


Male. — Reddish brown, declivity darker, 3.88 mm long, 3.33 
times as long as wide. A member of the group Crossotarsi 
subdepressi with the interstices 1, 3, 5, 7 costate on the upper 
convexity of the declivity. 

Front flat, shining, subimpunctate below, rugosely and longi- 
tudinally punctured above, with an elevated carina from the 
center towards the vertex, interrupted for a short piece above 
the center, upper surface rather densely hairy. 

Pronotum longer than wide (38 :33), lateral emarginations 
shallow; median sulcus short and very fine; surface shining, 
finely punctured ; punctures rather densely placed near the api- 
cal margin, scattered and inconspicuous behind. 

Elytra wider than pronotum (36 : 33) and 1.78 times as long, 
sides subparallel, broadly emarginate behind; disc very finely 
lineate-punctate, punctures hardly visible at the sides ; interstices 
flat and impunctate, somewhat convex near the base where the 

86. 8 Schedl: Fauna Philippinensis, III 399 

third is conjoined with the first and fifth, base of the third 
without any remarkable punctuation; declivity commencing in 
the posterior third, convex above, perpendicular and with a 
lunate depression below, the lateral processes bent backwards 
and downwards, the interstices 1, 3, 5, 7 costate on the anterior 
half of the upper convexity, nearly horizontal and then abruptly 
ceasing, the other interstices low, ending at the commencement 
of the carinse, the rest of the convexity opaque, all interstices 
continued as low ridges which are covered with a row of setose 
granules, the first spined at the upper border of the shining and 
impunctate lunate depression, the lower margin of the latter with 
a small tooth between the suture and the lateral process, the 
process with a slender spine at its apex. Abdomen normal. 

Female. — Colored as the male, 3.97 mm long, 3.26 times as 
long as wide. 

Front as in the male but a little more convex, punctuation 
and median carinse finer. 

Pronotum subquadrate (37 : 35) but of the same sculpture as 
in the male. 

Elsrtra but little wider than pronotum (36 : 35) and 1.80 times 
as long, sides parallel on the anterior half, narrowed behind, 
transverse at the apex; disc with sculpture similar to that in 
the male but still finer ; the base of the third interstice with a 
few minute granules; declivity opaque, densely hairy and feebly 
convex above, with a shining, lunate, perpendicular depression 
below; lateral processes blunt. Abdomen normal, convex. 

Types in the possession of Mr. F. C. Hadden and in my 

Locality. — Mount Maquiling, Laguna Province, Philippine Is- 
lands; F. C. Hadden, collector, 


Female. — Reddish brown, 4.17 mm long, 3.30 times as long as 
wide. A remarkable form, which is easily recognized by the 
characters of the front. I cannot place it in any group until 
the other sex is known. 

Front with a transverse, deep impression behind a narrow 
epistomal margin, the concavity extending up to the lower mar- 
gin of the eyes at the sides, much less so in the center ; a plush 
of long erect hairs on each side of the epistomal margin shortly 
in front of the antennal articulation; the concavity shining, the 
upper half of the front convex and densely covered with very 
long, erect, yellow hairs, abraded in one of the specimens; an- 

400 The Philippine Journal of Science 1935 

tennal scape about as long as wide, with a fringe of hairs on 
its outer margin, the longest of which are three times as long 
as the scape. 

Pronotum subquadrate (39 : 37), rather feebly convex, lateral 
emarginations rather shallow, surface shining, with three more 
or less distinct depressions on each side of the median line, 
median sulcus fine, surrounded by a circular patch of densely 
placed, fine punctures which rather gradually become coarser 
and less densely placed at the extreme margin; the remaining 
surface rather coarsely and moderately densely punctured and 
with a pubescence that is denser than in most Platypodidae and 
resembles that of Cenocephalus pusilVm Schedl. 

Elytra wider than pronotum (40 : 37) and 1.79 times as long; 
base finely carinate, sides parallel on the anterior two-thirds, 
narrowed along the upper convexity of the declivity, transverse 
behind; striate-sulcate, sulci indistinctly multipunctate ; inter- 
stices convex, the first narrow and with a row of tubercles on its 
entire length, the others with sparsely placed fine punctures, 
the punctures coarser and denser on the second, base of the third 
with a long row of crowded, round, fine granules; declivity 
convex, with granulate interspaces; shallow sulci and reddish 
pubescence above, with a perpendicular finely rugose face be- 
low, a plush of reddish hairs on the latter. Abdomen normal. 

Types in the Zoologisches Museum in Hamburg and in my 

Locality. — Philippine Islands. 


Male. — Black, 4.75 mm long, 3.37 times as long as wide. The 
third male known of the group Platypi coleoptrati. 

Front flat, hardly visibly impressed below, subshining, punc- 
tured on a triangular space below, coarsely, longitudinally 
wrinkled above, with a short, median, impressed striga. 

Pronotum 1.23 times as long as wide, shining, with a very 
long median sulcus, which is continued anteriorly by an im- 
pressed line, coarsely uniformly punctured all over. 

Elytra wider than pronotum (14.5 : 13.2) and 1.84 times as 
long, base carinate, disc striate-sulcate, sulci narrow, indistinct- 
ly punctured, all interstices equal, strongly convex, narrow, sub- 
impunctate; the ninth interstice densely irregularly punctured, 
near the base the sulci but little wider and more distinctly and 
densely punctured; declivity rather steep, truncate, subconvex, 
with three processes on each side, the apices of the median lower 
pair as widely separated from each other as the first process 

56' 3 Schedl: Fauna Philippinensis, III 401 

from the second, the third much closer to the second and formed 
by the carinate ninth interstice which terminates abruptly at 
the lateral margin of the declivital convexity, the first and fourth 
interstices decreasing in height and width towards the upper 
declivital margin, the others carinate and terminating abruptly 
at the commencement of the truncate declivity, the latter shin- 
ing and the interstices continued as rows of coarse granules. 
Abdomen normal. 

Type in my collection. 

Locality. — Philippine Islands. 


Male. — Nearly black, 4.26 mm long, 3.52 times as long as 
wide. The only relative of this species known may be C. taihei- 
zanemis Mur., from Formosa, which must be transferred to 
Stenoplatypus Strohm. It differs from the latter by the pres- 
ence of spines on the abdominal sternites and in the armature 
of the declivity. 

Front feebly convex, subopaque, with a few round punctures 
in front, longitudinally wrinkled behind. 

Pronotum slightly longer than wide (38 : 35), shining, me- 
dian sulcus hardly visible, with very fine and very sparsely 
placed punctures. 

Elytra wider than pronotum (37 : 35) and 1.86 times as long, 
striate-punctate, sutural striae impressed, others near the base 
and apex only; strial punctures indistinct to obsolescent; inter- 
stices convex towards the base and apex, with scattered very 
fine punctures, not remarkably stronger near the base; declivi- 
tal armature similar to that in C. taiheizanensis, but the fused 
first and second interstices forming a long blunt tooth, third and 
fourth subequal in length, the third broad and blunt, the fourth 
very narrow and without apical tooth; the lateral process con- 
sisting of a large blunt external and two small inner teeth ; the 
apical margin without armature; the third and fourth visible 
abdominal sternites with two widely separated pointed teeth, 
those of the fourth sternite being much larger. 

Female. — Nearly black, 4.36 mm long, 3.46 times as long as 
wide. A single specimen from an old collection and also origi- 
nating from Luzon, Philippine Islands, I am inclined to regard 
as the other sex. 

Front as in the male but with the sculpture shallower. 

Pronotum little longer than wide (40 :37), shining, with a 
long, strongly impressed, median sulcus and with three large 
pores, which are situated in an equilateral triangle at each side 

402 The Philippine Journal of Science isss 

of the sulcus, surface with widely scattered and very fine punc- 
tures, a small group of more densely placed punctures in front 
of the pores. 

Elytra wider (40 : 37) and 1.87 times as long as the prono- 
tum, sides parallel, narrowed along the declivital convexity, 
transverse at the apex; sculpture similar to that in the male, 
the striae becoming obsolescent behind; the base of the third 
interstice densely punctured and with some indications of low 
transverse ridges; declivity convex above, perpendicular below, 
rugose and with short yellow pubescence. Abdominal sternites 
convex and unarmed. 

Types in the possession of Mr. F. C. Hadden and in my col- 

Localities. — Males, Mount Maquiling, Laguna, Luzon, Philip- 
pine Islands, F. C. Hadden, collector; female, Luzon. 


Female. — Reddish brown, 2.37 mm long, twice as long as wide; 
closely allied to X punctulatm Egg. but stouter and much more 
densely pubescent. 

Front opaque, convex, indistinctly punctured except on the 
epiptomal margin where the punctures are deeper, finer, and 
more densely placed. Eyes strongly emarginate in front; an- 
tennal club distinctly wider than long. 

Pronotum globose, 1.26 times as wide as long, base transverse, 
posterolateral angles not rounded, rectangular; sides parallel 
on the posterior third, evenly broadly rounded in front ; anterior 
margin not armed ; summit behind the middle, anterior area fine- 
ly asperate, posterior area subshining, minutely reticulate and 
densely finely punctured. 

Elytra wider than pronotum (39.5 : 38) and 1.56 times as long, 
sides parallel, broadly rounded behind, convex from the base 
to the apex, apical margin not acute but rounded; the entire 
surface densely, finely, and irregularly punctured and with a 
fine, dense, yellow pubescence. 

Types in the possession of Mr. F. C. Hadden and in my col- 

Locality. — Mount Maquiling, Laguna Province, Luzon, Phil- 
ippine Islands, F. C. Hadden, collector. 


Female. — ^Reddish brown, eljrtra darker, 2.72 mm long, 3.34 
times as long a^ wide. A small species of the general shape of 
X. amphicranoides Hag. 

56, 3 Schedl: Fawm Philippinensis, III 403 

Front subshining, plano-convex, with a narrow, shallow, trans- 
verse depression between the eyes, which is interrupted by a 
low longitudinal elevation, finely, deeply, and densely punctured 
on the epistomal margin and towards the eyes, impunctate along 
the median line, more sparsely and shallowly punctured above. 

Pronotum shining, 1.20 times as long as wide, base trans- 
verse, posterolateral angles rounded, sides parallel on the an- 
terior two-thirds, rather broadly rounded in front; summit 
before the middle, anterior area densely finely asperate, pol- 
ished, finely and moderately densely punctured behind, impunc- 
tate along the median line. 

Elytra as wide and 1.75 times as long as the pronotum, sides, 
parallel on more than the anterior three-fifths, then rather 
obliquely narrowed, deeply emarginate at the suture; disc shin- 
ing, lineate-punctate, strise not impressed, consisting of regu- 
larly placed moderately fine punctures, those of the sutural 
strise somewhat coarser, interspaces flat and covered with an 
irregular row of finer punctures; dechvity commencing at the 
middle, concave as in X. amphicranoides, concavity subshining, 
subimpunctate, the suture elevated, lateral margin armed with 
two large slender spines, which are directed backwards, the first 
pair situated at the anterior fourth, the second pair at the 
middle of the declivity; second pair but little larger than th^ 
first, cephalad to the first spine with a pair of small sutural 
teeth and two pairs of small setose tubercles ; the apical emar- 
gination deeper than wide; the lateral processes with a small 
tooth, which is directed inwards on the inner margin, the outer 
one with a row of setose tubercles, the apices pointed. 

Types in the possession of Mr. F. C. Hadden and in my col- 
lection. The specimens bear the label: Ace. No. 195, Sch. of 
For. Univ. P. I. 

The Phiuppine 
Journal op Science 

Vol. 56 APRIL, 1935 No. 4 


Of the College of Agriculture, University of the Philippines, Los Banos 


Although many papers have been published on the "bunga" 
(Aeginetia indica Linnaeus), an important orobanchaceous root 
parasite ^ in the Philippines, it seemed desirable to undertake a 
study on the anatomy and morphology of the vegetative and 
reproductive organs of this plant because of the harm it is doing 
in sugar-cane fields in some parts of the Islands, especially 
Batangas and Laguna.(8,22,36,37) in Calamba Sugar Estate 
nearly a fourth of the total area devoted to sugar cane is more 
or less heavily infested with this parasite. Needless to say it 
is a serious problem and a source of no little worry to the admin- 
istration. Buiiga reduces the sugar content of the canes very 
considerably (36) and may even cause the formation of substances 
that render the canes unfit for milling. (22) The present paper 
is an attempt to describe those phases of the anatomy of the 
buiiga which form a background for a physiological study, and 
it is hoped this study will result in the development of a basic 
method of control. 

^College of Agriculture Experiment Station contribution, No. 1023. 
Read before the Los Banos Biological Club, July 26, 1934. 

'This plant is believed to possess medicinal properties by people in 
some towns of Batangas, where it is said to be cultivated to some extent. 

291458 405 

406 The Philippine Journal of Science i935 

According to Bakhuizen van den Brink, (3) this plant is now 
found in British East India, Java, New Guinea, Borneo, Phil- 
ippine Islands, Japan, and China, growing on roots of the 
Graminese, and a few of the Cyperacese, Zingiberacese, Cannacese, 
Gleicheniac^e, and Ericaceae. In the Philippines, bunga is dis- 
tributed in Luzon from Cagayan to Sorsogon, in all or most 
provinces, and in Leyte and Panay. It grows in these regions 
on various coarse grasses at medium altitudes. (35) In Rizal(22) 
and Negros, (8, 22, 37) as well as in Laguna, this parasite is mostly 
confined to sugar-cane plants.^ 

Kusano(28) traced the development of the primary tubercle 
and the primary roots of Aeginetia indica Linn."* in Japan, and 
he found that this parasite subsists generally on wild grasses 
(Miscanthus sinensis Anders, and Carex lanceolata Boott.) and 
a few cultivated plants [Zingiber mioga Rose, a mountain rice 
(Oryza sativa Linn.), Setaria italica Kth. var. germanica Trin., 
Zea mays Linn., Panicum miliaceum Linn., and Panicum fru- 
mentdceous Linn.]. His observation showed that its subter- 
ranean parts seem to winter in the vicinity of Tokyo in much 
the same way that it ratoons with the canes in the Philippines 
(Plate 9, fig. 90) . The seed, according to Kusano, upon germina- 
tion sends out first a filamentous radicle, its end swelling to 
form a primary tubercle, when proper host roots are present. 
Macroscopically he observed that the front of this tubercle 
penetrates into the cortex of the host root and becomes or- 
ganically connected with the latter, there differentiating into 
a primary haustorium. On the upper end the growing point 
of the stem develops. Simultaneously with the formation of the 
stem, numerous processes appear from the tubercle as roots of 
the parasite; these organs possess neither root hairs nor root 

* February 3, 1934, Mr. Emiliano F. Roldan, of the College of Agricul- 
ture, submitted for identification a grass (Paspalum scrohiculatum Linn.), 
which possessed young tubercles of the bunga. This grass, according to 
Mr. Roldan, was artificially inoculated in a pot of soil. The writer grew 
this same grass and several other graminaceous weeds, and these were 
inoculated with seeds of the bunga March 7, 1934. March 28, 1934, the 
following weeds showed the presence of young tubercles of various sizes 
from pinhead to about 5 millimeters in diameter: Eleusine indica (Linn.) 
Gaertn., Eragrostis amabilis (Linn.) W. and A,, and Paspalum scrohicula- 
tum Linn. 

•The English resume by the author was published by Roxas.(36) 

56.4 Jvliano: Bun§a 407 

He further stated that both stem and root are formed endoge- 

Later Kusano(29) performed extensive germination experi- 
ments on the seeds and he described in detail early stages of 
germination and development of the young seedling. According 
to this investigator the seeds did not germinate when immersed 
in chemicals, and direct contact with the host roots was by no 
means an indispensable condition in bringing them to germinate. 
Plants found to induce seedlings to form tubercles were : Luzida 
campestris DC. var. capitata Miq., Carex japonica Thunb. var. 
chlorostachys (Don) Kuk., C morrowi Boott., Miscanthits sinen- 
sis Anders., Calamagrostis arundinacea Roth., Setaria excurrens 
Miq., Oryza sativa Linn, (upland form), Panicum miliaceum 
Linn,, Pollia japonica Hernst., Zingiber mioga Rose, and Canna 
indica Linn. He also found that dry seeds lose viability in two 
years in Japan. 

Coert(i3) has described an Aeginetia^ (A. saccharicola 
Bakh.)^ plant from Java, which appears to be similar to the 
Philippine bunga, except in the number of its placentas. Into 
Coert's description were incorporated many useful observations 
of the habits and growth of this Javan parasite. 

McWhorter,(34) Teodoro,(46) and Roxas(36) attempted to lay 
out methods of control for the parasite; their recommendations 
were based largely on the findings of Kusano.(28,29) Despite 
all these prescribed methods of combating this plant and the 
presence of some natural means of control, its spread remains 
unchecked. If control measures are not found, it may be only 
a matter of time before its spread will be beyond human con- 
trol. Some investigators interested in its eradication have di- 
verted themselves with the use of herbicides, or weed killers. (37) 
Agati and Tand) used Atlacide in their attempts to find an 
effective means of killing this pernicious plant, but they found 
that their herbicide only helped to prevent seedage, the under- 
ground organs remaining unharmed. Later, the same investi- 
gators (2) applied DL solution in two concentrations and salt 
(NaCl) in three concentrations. Their results proved that these 

''A translation into English of the original article by the author ap- 
peared in Sugar News 9 (1928) 367-375. 
•See Bakhuizen van den Brink. (3) 

408 !rA.e Philippine Journal of Science 1935 

herbicides also were efficient in killing the aerial organs of the 
buiiga, but its underground parts remained unaffected. In fact, 
under large scale conditions, these authors observed that new 
sprouts appeared after the application of the spraysJ 


October 13, 1933, some of the material used in this study 
was obtained from the cultures inoculated June 12, 1933, by 
Mr. Emiliano F. Roldan. Later collections were made in the 
Calamba Sugar Estate ® as well as in the cane fields behind the 
Protestant Chapel, College of Agriculture, Laguna, in December, 
1933, and from January to May, 1934. Artificial germination 
of seeds in both black and white sand was made in the labora- 
tory, at the College of Agriculture. Before being inoculated, 
the cane seedlings (POJ 2878) were allowed to grow normally 
a week or two in the sand; then they were washed off carefully 
with water and their root systems were covered with the buiiga 
seeds collected in December, 1933, in the Calamba Sugar Estate, 
and with some obtained through the courtesy of Dr. G. 0. 
Ocfemia. From time to time individual cane seedlings were 
dug up to be examined for signs of germination and tubercle 
formation. In all cases the young cane seedlings were moder- 
ately fertilized with Nitrophoska. 

In cane fields adjoining the grounds of the Protestant Chapel, 
College of Agriculture, from February to July, 1934, periodi- 
cally, canes were observed and dug up. These observations 
were made prior to and after harvesting in order to determine 
the behavior of this plant parasite on ratoon canes as well as 
on newly planted cane seedlings. 

^An endemic root parasite (Christisonia wightii Elm.) very closely re- 
lated to Aeginetia indica Linn., was recently reported by Goseco (21) to 
infest sugar-cane varieties; namely, DI 52, POJ 2878, Badila, New Guinea 
24-A, and Negros Purple, in La Carlota, Occidental Negros. According 
to Elmer (16) his type specimen was collected on Cuernos de Negros, 
Dumaguete, Oriental Negros, March, 1908, growing in rich damp ground 
among the tussocks of Ammomum fusiforme Ridl. at an altitude of about 
1,750 feet. This parasite is also found in Laguna, Bataan, Negros, and 
Lanao, (35) growing at altitudes of from 400 to 800 meters. Attempts 
of the writer to grow under laboratory conditions this parasite from seeds 
obtained through the courtesy of Mr. Federico P. Goseco, of the Research 
Bureau of the Philippine Sugar Association, have failed. 

* Thanks are here expressed for the kind cooperation and help extended 
to the writer by the Calamba Sugar Estate in providing him with suit- 
able specimens of the parasite for this study. 

6«»4 Jvliano: Bunga 409 

The material, consisting of vegetative as well as reproductive 
organs, was fixed and killed in the laboratory with the use of 
formalin-acetic alcohol (70 per cent) and chromo-acetic stock 
solution prepared according to the formulae given by Chamber- 
lain. (11) It was treated in the usual fashion and embedded in 
paraffin. Sections 8 to 10 microns thick were cut and stained 
in Heidenhain's iron-alum hsematoxylin with aniline orange G 
dissolved in clove oil as a counterstain. Free-hand sections of 
mature vegetative organs were also made, and these were stained 
in either Heidenhain's iron-alum haematoxylin, Delafield's hsema- 
toxylin-safranin, or safranin-light green. 


Description. — This obligate parasite possesses numerous white 
underground roots producing numerous tubercles or swellings, 
which attach themselves to the host roots by means of sucking 
organs, or haustoria. The roots of this parasite may be de- 
scribed as white, irregularly ellipsoid, rhizomatous, brittle, and 
delicate underground organs of small diameters. They branch 
freely and grow rapidly with the meshes of the cane roots in 
a hill. In a buiiga seedling the primary roots are formed from 
its primary knoblike outgrowth, or primary tubercle, as was 
also observed by Kusano,(28) who found neither root caps nor 
root hairs developing from them. From a single tubercle, 
numerous roots as well as one to several fleshy, rather large or 
small, short, aerial, nonchlorophyllose scapes arise. The scapes, 
which may be 15 to 25 centimeters or more in height, bear few 
to many scalelike, rather erect, fleshy leaves; each scape is 
provided with few to many branches or stalks, each bearing a 
single terminal flower. The aerial organs, except the corolla,, 
pistil, stamens, and seeds, are purplish, longitudinally mottled 
by yellowish stripes. 

Anatomy. — ^A young root in transverse section (Plate 1, fig. 
1) shows an outer layer of parenchymatous, irregularly shaped, 
epidermal cells, the outer walls of which are often thickened and 
usually brownish, with granular matter attached to them. Below 
this single-layered epidermis is a rather extensive cortex of 
parenchjnnatous cells, each provided with an abundance of starch 
grains. These cortical cells are rounded to isodiametric in 
transverse section and tend to become smaller towards the cen- 
tral cylinder; in longitudinal section they are tubular, with 

410 The Philippine Jourrval of Science 1935 

their end walls set either horizontally or obliquely. The en- 
dodermis as well as the pericycle are not well differentiated. 
The phloem tissue of the central cylinder has just begun to 
differentiate. While the endodermis of Christisonia subacaulis 
Gardn.(52) as well as of Epiphegus virginianail^) is not easily 
distinguishable, that of Christisonia bicolor (Jardn. is distinct, 
although its pericycle is not clearly differentiated (52) as it is 
interrupted here and there by the phloem elements. In Aegi- 
netia indica Linn., however, the cortex of the young root can be 
easily separated from its central cylinder by the presence of 
starch grains in the former and their absence in the latter. 

The central cylinder of this young root is rather small, and 
it is devoid of starch grains as stated above. The primary 
phloem rays have already differentiated (Plate 1, fig. 1) , forming 
three distinct groups of small, isodiametric parenchyma. The 
ground tissue is largely composed of thin-walled parenchyma. 
The elements of the central cylinder are mostly elongated longi- 
tudinally with their long axes parallel to the length of the root, 
their end walls set either horizontally or obliquely. 

As the root matures, the triarch arrangement of the vascular 
bundles (Plate 1, fig. 2) becomes more evident. The primary 
xylem consists of reticulated tracheids with thickened walls, 
and the rays often meet at the center of the stele so that the 
pith is often practically nil. The tracheids are short and rec- 
tangular, and all are of about the same size. The primary 
phloem rays alternate with the primary xylem rays and are 
separated by parenchymatous tissue. Sieve tubes, companion 
cells, and parenchyma are found in the phloem. The primary 
phloem cells are all cylindrical in longitudinal section, rather 
long, and possess distinct nuclei, while their end walls are usually 
set horizontally, although obliquely set walls are not uncommon. 
The absence of vessels in the root is very evident; this very 
primitive condition of the vascular cylinder in which the xylem 
has become reduced and the phloem correspondingly developed 
exists also in Christisonia. (^^) 

As the root reaches full maturity, it becomes somewhat tough 
owing to the formation of a sclerenchymatous ring around its 
central cylinder (Plate 1, fig. 7). The epidermal cells become 
tangentially flattened in transverse section (Plate 4, fig. 28), 
their outer walls always remaining thick with granular matter 
covering them on the outside. The cortical cells are likewise 
fllattened and become somewhat spongy in character, assuming a 

56r4 Jvliano: Bunga 411 

variety of shapes. Some of these cortical cells may be rectan- 
gular, oblong, or otherwise constricted ; all contain an abundance 
of starch grains. The inner cortical cells of several layers in 
thickness are lignified, and these lignified cells may actually 
invaginate the central cylinder (Plate 1, fig. 7). Lignification 
usually takes place 2 or 3 centimeters from the apex of the root. 
Simultaneously with the completion of the thickening and ligni- 
fication of these inner cortical cells, their stored food disappears. 
Some of these cortical cells may be pitted; they are usually large 
and variously shaped. The presence of lignified tissue around 
the central cylinder of the root is reported also in Christisonia 
bicolor Gardn. by Worsdell.(52) 

The secondary phloem (Plate 1, fig. 4) consists mostly of 
parenchymatous cells with distinct companion cells and sieve 
tubes. The secondary xylem is composed of groups of reticu- 
lated tracheids, the walls of which are very much thickened. 
These groups of tracheids are often separated from each other 
by a row or two of distinct parenchyma. Between the secondary 
phloem and secondary xylem is an indistinct cambial layer, 
and this meristematic tissue does not form a complete ring 
around the central cylinder because of the invagination of the 
sclerenchymatous cortical cells. In fact, this cambium may be 
regarded as degenerated tissue as not much secondary thicken- 
ing takes place in the root. 

Occupying the central portion of the central cylinder are the 
primary xylem rays (Plate 1, fig. 7). 

Root apex. — ^The apex of the root (Plate 4, fig. 30) bears no 
root cap, as was reported also by Kusano,(28) being entirely 
covered by a distinct epidermis (dermatogen) . The following 
species are reported to lack root caps: Christisonia suhacaulis 
Gardn. and Ch. neilgherrica Gardn., (62) Epiphegus Virginia- 
na,(i4) and Cassytha filiformis Linn. (9) Root caps are present 
in Aphyllon uniflorum Gray, (44) Orobanche speciosa, 0. minor, 
O. ramosa, and O. hederseA^Q) The root apex of the parasite 
in question has the features of a typical monocotyledonous root 
without the root cap. The dermatogen consists of squarish to 
rectangular parenchymatous cells, the thin plerome, of rather 
much elongated, small, rectangular cells, and the periblem, of 
large, oblong to rectangular cells. The first covers the root, the 
second occupies the central portion, and the third represents 
the bulk of the root. All the cells of these embryonic tissues 
are provided with dense cytoplasm and distinctly large nuclei. 

412 The Philippine Journal of Science 1935 

At the apex of the root may often be found a mucilagelike sub- 
stance, which is perhaps essential in its growth through the 
soil and between the host roots. 

Origin of primary and secondary roots. — ^The endogenous 
origin of the primary as well as the lateral roots as reported by 
Kusano(28) is herein confirmed. The writer has observed that 
the development of the primary root from the primary tubercle 
takes place rather early ; that is, the root initials may be seen to 
differentiate in the tubercle even before the latter has attained 
a size of 0.5 millimeter. The primary root is first recognized, 
while still deep in the thin cortex of the tubercle, as a group 
of embryonic cells (Plate 1, fig. 8). Because of the lack of 
clear demarcation between the cortex and the central cylinder 
in the tubercle, the writer was not able to determine with ab- 
solute accuracy the particular tissue in the tubercle that gives 
rise to the primary root. However, these embryonic cells push 
themselves through the parench3miatous and starchy ground 
tissue of the cortex of the tubercle, and as this root initial ad- 
vances outward, those cortical cells behind it gradually form the 
vascular bundle, which later becomes the central vascular bun- 
dle of the primary root. The front of the advancing embryonic 
cells, on the other hand, forms a distinct row of epidermal cells 
upon reaching the circumference of the tubercle (Plate 1, fig. 
3). Traces of distorted and destroyed cortical cells may be 
seen along the front of the advancing primary root. 

Development of the lateral roots in Christisonia bicolor Gardn. 
and Ch. subacaulis Gardn., (52) as well as in Epiphegm virgi- 
niana,(W is closely similar to that in Aeginetia. The lateral 
roots of these plants are first differentiated as a group of em- 
bryonic cells pushing out from one of the undifferentiated 
central bundles. (14) These cells advance through the tubercle, 
leaving behind a gradually forming bundle that becomes the 
central bundle of the lateral root. By the time this advancing 
group of embryonic cells has reached the periphery of the tu- 
bercle, it has formed a distinct row of epidermal cells along its 
apex. In some such roots the cone-shaped apex may still pos- 
sess a considerable area of embryonic cells containing large 
nuclei and densely stained cytoplasm. Worsdell(52) states that 
most probably the pericycle forms the lateral roots in Christiso- 
nia subacaulis Gardn. 

Development of secondary or branch roots in Aeginetia takes 
place very near the apex at right angles to the mother root. 

56, 4 Juliano : Bunga 413 

The first indication of the development of a branch root is shown 
by the activities of the cortical cells at the region between two 
adjacent phloem rays (Plate 8, fig. 83). These cortical cells 
acquire denser cytoplasm and distinctly stained nuclei compared 
with their surrounding cells, and these differentiating cortical 
cells become somewhat actively engaged in division at all planes 
so that they form a distinct strip of meristematic cells. Later, 
organization of the embryonic tissues of the young branch root 
takes place (Plate 8, fig. 84), and as this young branch root 
pushes itself through the cortical cells of the mother root, those 
cells left behind form the vascular bundles of the young branch 
root, similar to that obtaining in the primary root (Plate 1, fig. 
3). As this root emerges from the cortex of the mother root, 
it forms a distinct epidermis on its front. 

General consideration. — ^The tracheids of the stele and the 
sclerenchymatous cylinder of the cortex in the mature root stain 
red with phloroglucin-hydrochloric acid. When roots are ex- 
posed to the air, the cortex readily loses its water content, and 
darkening of the root takes place at once. This darkening is 
perhaps due to the abundant tannin in the cell sap of the cor- 
tical cells. Storage of starch grains takes place mainly in the 
cortex, a tissue which easily allows the escape of moisture upon 
exposure to air, while the sclerenchymatous cortical cylinder 
shuts off the central cylinder from the food supply stored in 
the mature root. 

The cellulose nature of the outer walls of the epidermis as 
well as the cortical tissue of the root does not preclude the 
possibility that water absorption may not take place at all 
through this underground organ. Intake of water with dis- 
solved mineral matter from the soil may be possible only at 
regions where the cortical sclerenchymatous cylinder has not 
yet fully developed, and this usually takes place far from its 
apex. In fact, Cooke and Schively(i4) were even tempted to 
state that the roots of Epiphegus virginiana are also perhaps 
for the absorption of water. 

In numerous excavations made by the writer in cane fields 
planted to POJ 2878 behind the Protestant Chapel, College of 
Agriculture, Laguna, from February to June, 1934, where this 
buiiga was producing only one to two clusters of fresh flowers 
in late February, the underground organs of this parasite were 
found to be still fresh and connected with the cane roots. As 
the fruit matures, the main axis of the scape becomes dry, and 

414 The Philippine Journal of Science i»35 

breaks oflF early from the fresh underground tubercle. Ter- 
mites also help in mechanically detaching the scapes from the 
tubercles. When the canes were harvested in March, 1934, these 
aSrial parts of the bunga were mechanically destroyed and their 
seeds disseminated. Some of the seeds and remains of the aerial 
organs might have been destroyed by fire when the trash was 
burned. In fields where the ratoon canes were allowed to grow, 
excavations made April 28, 1934, showed that the roots of the 
old bunga were still fresh and were attached to the cane roots 
(Plate 9, fig. 90), despite the fact that the fire had cleared the 
field of rubbish. In several instances the writer observed that 
even ordinary secondary roots of the parasite may start to 
develop scapes without first forming tubercles. What was of 
greater import to the writer was the development of a scape at 
the injured apex of the root (Plate 9, fig. 91, A). It seems 
quite possible that the bunga was stimulated somewhat by the 
heat of the fire to produce an abundance of reproductive organs. 
In other words, scapes are not only produced from either prim- 
ary or secondary tubercles (Plate 9, fig. 91, B), as reported by 
Kusano, (28) but these reproductive organs may also be formed 
by ordinary roots of the parasite, especially in ratoon canes. 
Later visits to the same field showed that the bunga had actually 
become biennial (Plate 10, fig. 92) , and perhaps would become 
perennial if left on the same ratoon cane for a number of years. 
This same behavior of the parasite has been observed in the 
Calamba Sugar Estate on ratoon canes (PSA 14 and POJ 2878) .• 
It is apparent that ratooning canes are favorable to the peren- 
nial growth of this parasite. 

Another field just as heavily infested with buiiga as the one 
specially observed and adjacent to these ratoon canes, was 
planted to POJ 2878 canes April 15, 1934. The soil was well 
prepared, and excavations made by the writer about the middle 
of May, when the canes were 2 to 3 feet high, and June 14, 1934, 
when they were 4 feet high, showed no sign of infection, indi- 
cating that perhaps the old parasite had been completely killed, 
and infection must of necessity have come from old seeds. 

From the foregoing anatomical characteristics of the under- 
ground organs of this parasite, the writer is inclined to believe 
that infestation of new seedling canes under field conditions by 
chii^ of roots detached from the host is not only improbable 

•From the monthly report of the Calamba Sugar Estate Research De- 
partment for June, 1934. 

66,4 Jvliano: Bunga 415 

but impossible. Infestation of these new cane seedlings must 
be attributed to seeds left by previous crops of the parasite. 
Besides, the ease with which moisture actually escapes from its 
roots makes it less difficult to prevent further growth of the 
root pieces, especially if exposed to air prior to their being 
carried to the roots of the new cane seedlings. In fact, the 
time needed to prepare the land for the next crop of cane, and 
the rigorous method of cultivation followed prior to planting 
the cane points, are enough to desiccate the roots of the para- 
site. True, comparatively few seeds actually germinate in 
nature, as a great number of them are germless, but only one 
seed is necessary to infect a hill of canes as the young seedling 
parasite can vegetate quite successfully (Plate 10, fig. 93). 

It is highly probable that a tubercle or even a segment of the 
parasite may continue to produce roots and reproductive organs 
when perchance it is mechanically severed from its parent para- 
site, provided this severance is preceded by the establishment 
of organic connection between the severed segment and the host 
root — Si point that needs further study. Otherwise, the detached 
part or segment will die. Infestation of nearby hills may take 
place if the canes are planted so near each other as to allow 
their root systems to overlap. In fact this parasite can grow at 
some distance from the hill. 

In connection with the infestation of new cane seedlings by 
a mature parasite, the possibility of infesting new cane plants 
with old stumps of the parasite left attached to the old host 
canes must be admitted, especially under controlled laboratory 
conditions, where factors are made so favorable for growth. 
To test this possibility one would have to eliminate all the seeds 
clinging to the parasite and the cane stumps. This means of 
infestation, if possible, is, in the writer's opinion, analogous to, 
if not identical with, conditions obtaining in ratoon canes already 
infested with the parasite. While the parasite that is made 
to infest new canes can hibernate or even continue to produce 
new tubercles and reproductive organs as it has an abundance 
of stored food and organic connection with the old host roots is 
not destroyed, it can transfer to the new cane seedling in much 
the same way as the parasite on ratoon canes and renew its 
growth on roots arising from the developing buds of the old 
ratoon canes in the field. This phase of botanical inquiry, 
which awaits verification, has some important significance in 
the control of the parasite. 

416 The Philippine Journal of Science i9S5 


Kusano(29) described in detail the development of a primary 
tubercle soon after germination of the seed. According to him, 
after the hair-tendril cells are formed by the germinating seed, 
the cells below this filamentous radicle form a spherical or oval 
primary tubercle by rapid cell multiplication. This parenchy- 
matous tissue of the young tubercle pushes and breaks the hair- 
tendril cells, and comes to lie in contact with the host root. 
Until organic connection is made, cell multiplication within 
the tubercle is due to the reserve food in the parasite, and 
Kusano claims that the young tubercle may attain the full size 
of about a millimeter with the help of its endosperm. 

The anatomical structure of the tubercle as found by Ku- 
sano (28) is herein given. The tubercle possesses numerous vas- 
cular bundles of a peculiar appearance, and these are scattered 
on the ground parenchyma filled with an abundance of starch 
grains. Surrounding the tubercle is an epidermal layer of cells 
which may be broken at intervals, and these epidermal cells 
have their outer walls heavily stained by hsematoxylin. At 
regions where the epidermis is broken or discontinuous, the 
outer cortical cells actually function as the epidermis. Below 
the epidermis is the cortex consisting of large, parenchymatous 
cells provided with an abundance of starch grains. This cortex 
is continuous with that of the main root and usually incloses 
the region containing the vascular bundles (Plate 1, figs. 5 and 6 ; 
Plate 2, fig. 15). 

Scattered on the ground parenchyma of the central portion 
of the tubercle are vascular bundles. In the middle of the 
individual bundle (Plate 2, fig. 11) are few to numerous lignified, 
thick-walled tracheids with reticulate thickenings. Surrounding 
these tracheids are one to several layers of small, more or less 
elongated, parenchymatous cells, with very dense cytoplasmic 
content and rather large nuclei. Separating the individual small 
or large bundles are few to several large layers of other paren- 
chymatous cells, with clear nuclei and not as dense cytoplasm as 
those found directly surrounding the tracheids. Two or more 
vascular bundles may fuse on entering a haustorium, while their 
branches diverge towards the central cylinder in the mother 
root. A direct connection with the central cylinder cannot be 
discerned ; their ends appear to dip away into the cortex. The 
absence of a sclerenchymatous ring of cells around the central 
cylinder is manifest. The formation of such prominent and 

56,4 Jvliano: Bunga 417 

peculiarly disposed vascular bundles has also been described in 
the tubercle of Christisonia subacauUs Gardn. by Worsdell,(52) 
who believes that their presence in the tubercle affords a good 
way of distributing the food absorbed from the host to points 
where shoot and root formation is very active. 

The formation of the primary tubercle is relegated to the 
seedling, but that of the secondary tubercles, to the primary as 
well as to branch roots, and the structure of the secondary tu- 
bercle is similar to the primary tubercle. The development of 
the secondary tubercle is macroscopically indicated by a distinct 
enlargement of the secondary or primary roots — an enlarge- 
ment which is often asymmetrical at first, becoming rounded 
later. Internally the cortical cells of the root of the parasite, 
adjacent to the host root, exhibit very active divisions in all 
planes (Plate 2, fig. 10; Plate 8, fig. 85), so that the secondary 
root often reveals a heterogeneous cortical tissue. The cortical 
cells away from the host root are large, rounded, and more or 
less spongy in character, while those facing the host root are 
much smaller and possess distinct nuclei and denser cytoplasm. 
In the latter cells, starch grains are not abundant, and this is 
perhaps due to their rapid consumption during cell divisions. 
Later, some of the cortical cells divide radially, so that rows 
of radially elongated cells are formed, which become the fore- 
runners of the vascular bundles in the young developing tuber- 
cle (Plate 8, fig. 86). 


How connection with the host root is effected in an Aeginetia 
seedling is described by Kusano. (29) He found that upon germi- 
nation the seed sends out from its radicular end two or three 
large, hyaline, globular cells, which are highly turgescent, and 
each is provided with abundant cell sap. Their nuclei are large 
and conspicuous, and the cytoplasm radiates from them. Those 
globular cells increase in number generally up to fifteen, ap- 
proximately, comprising the epidermal cells of the radicular end 
of the embryo, swelling to nearly four times their original dia- 
meter. Kusano claims it is improbable that multiplication of 
cells may be partly concerned in the increase in size. Each 
globular cell protrudes its external wall, which has first a conical 
and then a papillalike form. These outgrowths proceed further 
to form slender hairs, some of which grow to a length of about 
a millimeter. These hairs are either septate or even branched, 
and are rhizoidal in character. They usually grow straight and 

418 The Philippine Journal of Science i9S5 

radiate in all directions if undisturbed ; if they come in contact 
with the host root, they seem to attach themselves firmly to it, 
and then coil or contract through their whole len^h, whereby 
the seedling is drawn closer to the host. Kusano was not able 
to determine with absolute accuracy how the hairs fix themselves 
on the host root. The hypodermal cells below the hair-tendril 
cells of the embryo form the tubercle, which penetrates the host 
root, and thus becomes organically connected with the host root 
by means of a haustorium. 

In Orobanchei21) the radicular half of the embryo develops 
into a filamentous radicle, the plumular half remaining through- 
out in the endosperm acting as an absorbing agent. Parasitism 
here is effected by the root tip. 

The seed of Phelipaea ramosa,{iO) upon germination, sends 
out a long, filamentous, multicellular radicle devoid: of root 
hairs, the end of which is enlarged. This radicle pushes itself 
through the endosperm, while the chalazal end of the embryo 
gives rise to the stem. The tip of the radicle penetrates the 
cortex of the host root. Vascular bundles of the root are 
developed, but no connection can be seen between its main bundle 
and the haustorium. The upper end of the haustorium enlarges 
and later sends out adventitious roots. 

Massee(33) pointed out that the seed of Lathraea squamaria 
Linn, sends out a tap root upon germination, and this root gives 
off numerous branches, each furnished with minute hemispher- 
ical suckers or haustoria. The root forms short root hairs and 
these penetrate between the epidermal cells of the host, while 
the central vascular portion penetrates deeply until it reaches 
the pericycle of the host from where it absorbs nutriment. 

How organic connection is made by Melampyrum pratense 
with its host has been fully described by Leclerc du Sablon.(30) 
The two-layered cortex of the root greatly elongates at the 
region facing the host root, and the cells there divide in various 
directions, forming a swelling; the epidermal cells, which were 
first isodiametric, send out lateral protuberances. One of the 
epidermal cells towards the host elongates tangentially at first, 
and then divides radially or anticlinally several times, some of 
its daughter cells actually elongating. These elongating cells 
form rootlike outgrowths, which penetrate the host root. The 
pericycle does not in any way remain inactive and its cells 
elongate and later divide periclinally at the region contiguous 
to the enlarged cortical region, simulating the formation of a 

56.4 Jvliano: Bunga 419 

branch root; this forms the major portion of the haustorium. 
This activity of the pericycle presses down the endodermis so 
that it actually disappears at this region of the root. The 
piliferous layer may, in certain cases, exhibit very active radial 
divisions, the daughter cells of which elongate to form a compact 
mass of hairs. These hairs also penetrate the host roots. Hairs 
formed by the epidermal cells grow into the host much as the 
mycelia of a fungus penetrate its host, dissolving the host cells 
on their way by simple diastase reaction. These hairs form a 
**vascular system" with spiral thickenings, and differentiation 
of their reticulations proceeds until there is formed a continuous 
canal from the hairs to the central cylinder of the main root. 
A similar development of the haustorium has been described by 
this investigator for Melampyrum cristatum, M. silvaticum, and 
M. nemorosum. In Tozzia alpina, as well as in Rhinanthvs 
(Alectorolophics) and Pedicularis, Leclerc du Sablon(30) believes 
the haustorial development is identical with that of Melampyrum. 
Barber(4,5,6) is of the opinion that the formation of haustoria 
in Olax scandens and Santalum album is similar to that given 
by Leclerc du Sablon(30) for Rhinanthacese (Scrophulariaceae). 

In Christisonia bicolor Gardn.(52) the external cells of the 
root at the point of contact divide in places, and the cells for 
some distance on either side become elongated radially, while 
they become filled with dense cytoplasmic content and exhibit 
conspicuous nuclei; at the same time cortical cells immediately 
below this layer divide rapidly in succession in several directions; 
a few divisions occur in those more deeply situated. As a 
result of those cell divisions in the initial stage of haustorial 
formation, the cortex bulges considerably. A few cells on either 
side of the point of contact grow into hairlike papillae, which 
possess thick walls and dark brown contents with prominent 
nuclei. These grow towards the host root. The portion of the 
cortex at the point of contact at length, by repeated divisions in 
the cells, grows out as the haustorium, the cells in contact being 
much elongated and more or less contorted, with dense cjrto- 
plasm. Its origin is, therefore, exogenous, and no sieve tubes 
are present. Its development is quite different from that of 
Rhinanthus and Curscuta. 

In Aeginetia indica Linn, the writer believes entrance into 
the host root is effected by the pressure exerted by the enlarging 
epidermal and cortical cells of the tubercle facing the host root 
(Plate 2, fig. 13). The haustorium then penetrates the host 

420 The Philippine Journal of Science nz^ 

cortex through the endodermis into the stele of the sugar-cane 
root (Plate 2, fig. 15) . The ground tissue of the haustorium is 
parenchymatous, and its cortex is continuous with that of the 
tubercle. The cells (Plate 2, fig. 12) that abut the vessels of 
the host root are variously shaped, and contain dense cytoplasm. 
These cells become somewhat tubular to rectangular upward in 
the haustorium. Some of the central cells of the haustorium 
may become reticulately thickened and form the conductive 
system connecting the parasite with the vascular bundles of the 
host root. Connection is made not only with the xylem vessels 
but also with the phloem tissue of the host root, so that the 
parasite actually draws water and elaborated food from the host. 
The haustorial bundle branches into two or more divergent 
bundles before passing into the central cylinder of the tubercle, 
and these bundles may oftentimes dip away into the cortex 
without exhibiting actual connection with vascular bundles of 
the main root. Worsdell(52) claims that in Christisonia these 
bundles act as storehouses for the spoils of the haustorium, and 
by traversing the enlarged cortical tissue in all directions aid 
greatly in distributing the nutriment throughout the tissues of 
the tubercle so formed, supplying the young lateral roots and 
scapes with an abundance of food. 


General consideration. — The scape arises endogenously(28) 
from the tubercle. Under laboratory conditions, scapes are 
formed about two to three months after inoculation of the cane 
seedlings in sand. In the formation of scapes the tubercle 
(Plate 9, fig. 91, B) is the one organ responsible for their devel- 
opment, but in ratoon canes the writer has observed that scapes 
may arise directly from the main root of the parasite without 
first forming a tubercle (Plate 9, fig. 91, A). 

Under field conditions the bungra usually flowers more or less 
simultaneously with the sugar cane or oftener ahead of it. 
Even after the canes are harvested, the buiiga may continue 
to produce reproductive organs (Plate 9, fig. 90), perhaps to tide 
them over until the ratoon canes have started to grow again. 
Actually, however, the parasite starts to form scapes from its 
tubercles long before the flowering season of the sugar cane 
comes. Simultaneously with the development of the scapes 
and their emergence from the soil, the sugar-cane plants form a 
thick covering over the area that conceals the flower buds of 

56,4 Jvlimto: Bunga 421 

the parasite. When the canes are so tall that their basal leaves 
have dried up, a clearer view of the basal portion of the canes 
is obtainable, and at this time the bunga flowers are maturing. 
It is only at this stage of development that the bunga is actually 
recognized in the field, and by this time the damage to the 
canes has been done, and buiiga-seed production is perhaps at 
its best. As the bunga is an ombrophyte, its aerial organs are 
less vigorous and may even degenerate when this parasite is 
growing under direct sunlight. Consequently it does better in 
the shade, a condition most favorable also to the growth of the 
Javan Aeginetia saccharicola Bakh., which was reported by 

The scape is colorless before its emergence from the surface 
of the ground, and forms its peculiar pigmentation on exposure 
to air. It is short, terete, somewhat rigid, atropurpureous, 1 to 5 
centimeters, rarely up to 15 centimeters, long, and from 0.3 to 1 
centimeter in diameter. 

Origin. — ^The endogenous origin of the scape as reported by 
Kusano(28) is correct. A group of embryonic cells is differen- 
tiated deep in the cortex of the tubercle (Plate 8, fig. 87) and as 
these cells grow outward through the cortex, the parenchyma- 
tous cells behind them form vascular bundles to the scape. 
These embryonic cells, as they approach the periphery of the 
tubercle, cut off lateral protuberances (Plate 8, figs. 87 and 88), 
which form the scale leaves acting as protective covering to the 
growing point of the scape (Plate 2, fig. 14; Plate 8, fig. 89). 

Anatomy of the main axis. — A transverse section of the cen- 
tral axis to the scape, taken at or near its attachment to the 
tubercle, reveals the presence of bicoUateral vascular bundles 
arranged in a ring on a ground parenchyma filled with an abun- 
dance of starch grains (Plate 2, fig. 16; Plate 4, fig. 33). It is 
somewhat circular in outline, and is surrounded by a distinct 
epidermis of small, oblong to rectangular or squarish cells, the 
outer walls of which are somewhat papillate. In the sap of 
the epidermal cells is dissolved the pigment to the scape. In- 
closed by this epidermis is a thin cortex of small or large, 
oblong to rounded, tubular, parenchymatous cells, which are 
loosely packed together, and each is provided with a copious 
supply of starch grains. These cortical cells are smaller near 
the epidermis and become progressively larger towards the 

291468 2 

422 The Philippine Journal of Science 1935 

stele. The endodermis, or starch sheath, and the pericycle are 
not differentiated in this organ, as is true of Christisonia sub- 
acaulis Gardn.(52) 

The individual vascular bundle (Plate 2, fig. 16) is more or 
less rhomboid, and is inclosed by a bundle sheath consisting of 
three or four layers of sclerenchymatous cells. Surrounded by 
the sclerenchymatous bundle sheath is the phloem, which consists 
of distinct companion cells, sieve tubes, and parenchyma. The 
interior and exterior phloems have the same structure. The 
few tracheids, which occur in groups, are arranged in small 
broken rings and have thickened walls and reticulated thicken- 
ings. These surround a group of parenchyma at the center of 
the bundle. Separating the groups of tracheids and the phloems 
is an inner and outer discontinuous cambiumlike tissue one to 
two layers thick. 

The bicollateral bundles are separated from each other (Plate 
2, fig. 16; Plate 4, fig. 33) by few to several layers of cells sim- 
ilar to those found in the cortex; these are filled with starch 

The pith is made up wholly of large, parenchymatous cells 
with a copious supply of starch grains and mostly rounded 
to isodiametric in transverse section, similar to those present 
in the cortex. 

The number of vascular bundles of the main axis increases 
upward as a result of branching, and this is correlated with an 
increase in diameter. The epidermis and the cortex remain as 
before, while the pith is somewhat enlarged and possesses the 
same characteristic cell structure as described above. The bi- 
collateral bundles are more numerous and lie closer together 
(Plate 3, fig. 17), forming at times an almost continuous 
ring owing to the fusion of their sclerenchymatous bundle 
sheaths. In the majority of these bundles the sclerenchymatous 
sheath actually invaginates radially into the bundle from its 
outer tangential side, thus forming a U-shaped bundle (Plate 3, 
fig. 17). 

The individual bundle is also bicollateral (Plate 3, figs. 17 to 
19). The tracheids are arranged in groups and separated by 
parenchyma. On the outer and inner tangential sides are the 
phloem groups which consist of sieve tubes, companion cells, and 
parenchyma. Separating the tracheid groups and the phloems 
are scattered cambiumlike cells. 

56,4 JuLiano: Bunga 423 

WorsdeU(52) observes the presence of a ground tissue of hex- 
agonal to rounded cells filled with starch grains, and the vascular 
bundles are arranged in an irregular ring in the stem of Chris- 
tisonia. These bundles lie side by side in groups, separated by 
wide spaces of ground tissue. The individual bundle is circular 
in outline and is concentric in Christisonia neilgherHca Gardn. 
In stems of Aphyllon uniflorum Gray, (44) as well as in Epiphegus 
virginian(i,(W bicoUateral vascular bundles are present; these 
stems possess an abundance of starch grains, which pass into 
the soil with the decay of the plant. 

The main axis is devoid of stomata and trichomes, and is 
entirely glabrous. In Christisonia neilgherrica Gardn. stomata 
are present, rising above the surface of the stem, while in 
Ch. subacaulis Gardn. and Ch. bicolor Gardn. stomata are en- 
tirely absent. (52) Cooke and Schively(i4) attribute the wither- 
ing of Epiphegus virginiana after several of them have attacked 
a host, to the presence of numerous stomata. In Aeginetia in- 
dica Linn, nonresistance to the desiccating effect of direct sun- 
light is due to the absence of a heavy protective coating on the 
surface of the scape. 

At maturity of the fruits, the main axis of the scape decays 
and actually becomes separated from the tubercle that bears 
it. Decay of the main axis begins at its epidermis, and proceeds 
into the cortex and into the pith at times; only the skeleton of 
the vascular bundles is left to hold the scape erect. At a slight 
mechanical pressure this scape becomes automatically severed 
from the parent parasite, and thus helps in disseminating the 

Scale leaves. — The main axis of the scape bears a few to 
several scale leaves, which are spirally arranged on it (Plate 3, 
fig. 20) . Each of these scale leaves subtends at its axil a branch 
or pedicel, which carries at its apex a single flower (Plate 5, fig. 
34) . At maturity each scale leaf is more or less glabrous, some- 
what erect at first, later becoming horizontal, oblong ovate, 
small, and acuminate, acute or even obtuse, its base more or less 
clasping the main axis. It is about 0.5 to 2 centimeters long, 
and 0.3 to 0.5 centimeter broad. Because of the elongation of 
the main axis above a scale leaf, the pedicel or branch subtended 
by the scale leaf is carried way up, so that at maturity the 
scale leaf is borne far away from the pedicel or branch that 
it subtends. 

424 The Philippine Journal of Science i93& 

In transverse section the scale leaf is lunar in shape (Plate 
3, %. 20) and possesses distinct epidermal layers consisting 
of small, laterally compressed, rectangular cells on the dorsal or 
nether surface, and large, rounded to isodiametric cells with pap- 
illate outer tangential walls on the ventral or upper side (Plate 
1, fig. 9) . In these epidermal cells the pigment of the scale leaf 
is found. These epidermal cells are all elongated in surface 
view with their long axes parallel to the length of the scale leaf, 
their end walls being mostly set horizontally, although a few 
of them may be set obliquely. The mesophyll is occupied by 
loosely packed, parenchymatous cells, which are isodiametric; 
in this mesophyll are distributed the vascular bundles. These 
vascular bundles, which are poorly developed, lie near the ventral 
or upper side of the scale leaf. A single vascular bundle enters 
a single scale leaf, and this single bundle branches out several 
times as it traverses the whole length of the mesophyll of the 
scale leaf. 

On the upper side of the scale leaf are often a few, well- 
developed stomata consisting of two very much elongated, bean- 
shaped, guard cells, with their long axes parallel to the length 
of the scale leaf. The first organ of the parasite to show signs 
of desiccation is the scale leaf, and this susceptibility to drying 
is due perhaps to the presence of few of those natural openings. 
On the scale leaves of Aphyllon uniflorum Gray (44) stomata are 
found on the undersurface, while in those of Christisonia sub- 
a^aulis Gardn.(52) stomata and mucilage-secreting glands are 
found scattered on their ventral sides. 

Stalk or pedicel — ^Subtended by a scale leaf is a stalk or pedicel 
(Plate 3, fig. 20), which holds a single flower at its apex (Plate 
5, fig. 34). It is more or less erect, terete, and thickened up- 
ward, glabrous, atropurpureous, 10 to 30 centimeters long, and 
0.1 to 0.4 centimeter in diameter. 

In transverse section the young pedicel shows the presence of 
distinct groups of procambial strands (Plate 3, fig. 21) arranged 
in a ring on a ground of parenchymatous tissue. The epidermis 
consists of radially elongated cells, in which is found the pig- 
ment, and these epidermal cells inclose a rather wide cortex of 
thin-walled cells. The cortical cells are rounded to oblong and 
larger than the epidermal cells. The pith cells are similar to 
those found in the cortex and this ground parenchyma is pro- 
vided with an abundance of starch grains. 

®«'^ Jtdiano: Bunga 425 

As the stalk or pedicel matures, its epidermal cells become 
tangentially flattened, and are rectangular or more or less elon- 
gated vertically. The cortex becomes proportionately smaller 
and thinner, and consists of several layers of large, rounded to 
isodiametric, parenchymatous cells in transverse section, between 
which are small intercellular spaces. These cortical cells con- 
tain starch grains and tend to become larger from the epidermis 
inward to the middle portion, and gradually become smaller 
again towards the stele. The endodermis is indistinct or per- 
haps absent. 

The individual vascular bundles (Plate 4, fig. 29) are collat- 
eral, and arranged in a ring. The phloem, which forms the 
major portion of a bundle, consists of distinct parenchyma, com- 
panion cells, and sieve tubes. Running radially through the 
phloem are uniseriate or biseriate rows of hexagonal ray cells. 
The sieve tubes and companion cells occur in groups. The xy- 
lem, which lies towards the pith, is composed of groups of re- 
ticulated tracheids with thickened walls and separated by 
parenchyma. Between the tracheids are also found the hexag- 
onal ray cells. The xylem and the phloem are usually sepa- 
rated by distinct cambium consisting of one or two layers of 
cells occurring in patches. 

The pith is occupied by large, rounded, parenchymatous cells 
rich in starch grains and intercellular spaces. 

The mechanical tissues to the pedicel consist of the sclerified 
and lignified inner cortical cells (Plate 4, fig. 29) , as well as the 
lignified groups of cells separating the individual bundles and 
the peripheral pith cells. All these lignified tissues contribute 
to the rigidity of the pedicel. 

At maturity of the fruit, the pedicel turns black first at or 
near the point of its attachment to the capsule, and this blacken- 
ing progresses downward to the main axis of the scape. The 
pith cells are resorbed, and at the time the pedicel is dry, the 
pith is hollow. With the decay of the pedicel the unused starch 
grains in the cortex pass into the soil. 

Stomata are absent on the pedicel of Aeginetia indica Linn. ; 
they are present in ConophoUs americanai^i) and Aphyllon unu 
florum Gray. (44) In ConophoUs there are two concentric rings 
of separate collateral vascular bundles, the respective phloems 
of both rings facing each other. 

426 2^'^^ Philippine Journal of Science 1935 

The flower. — Each scape develops from one to a dozen flowers 
(Plate 5, fig. 34), each of which is long-pedicellate. Its calyx 
is ovoid, compressed, its base inflated, apex acute, more or less 
apiculate, coriaceous, glabrous, and very much shorter than the 
corolla. Externally it is purplish with longitudinal yellow 
stripes ; internally it is white or pale yellow, becoming blackish 
purple upward on its dorsal side. It measures from 1.5 to 3 
centimeters, rarely as much as 5 centimeters, in length and 1 
to 1.5 centimeters in diameter. Inclosed by it is a colorless, 
mucilaginous liquid secreted by multicellular glandular hairs 
(Plate 5, figs. 38 and 39) on its inner surface. 

Anatomically the calyx (Plate 3, fig. 22) has a distinct outer 
and an inner epidermal layer. The outer epidermal cells are 
much smaller than the inner, and their outer tangential walls 
are papillate. On surface view the outer epidermal cells are 
oblong to irregularly trapezoidal or nearly isodiametric with 
straight walls. The inner epidermal cells are much larger than 
those from its outer epidermis, more or less rectangular to 
oblong or even rounded in shape, their outer walls somewhat 
papillate, but almost always straight. On surface view these 
are identical with those of the outer epidermis of the calyx, but 
above the vascular bundles the epidermal cells are somewhat 
elongate. Both epidermal layers bear the pigment of the calyx. 
On the inner epidermis are borne, here and there, numerous 
glandular hairs, or trichomes (Plate 5, figs. 38 and 39), con- 
sisting of a uniseriate, a few- to several-celled stalk and a multi- 
cellular, glandular apex, the cells of which are often pigmented. 
In Epiphegus virginianaiW the trichomes are on the outer por- 
tion of the calyx and absent from the inside, while in Christiso- 
nia bicolor Gardn.(52) the trichomes are borne inside. Stomata 
are absent from the calyx of Aeginetia indica Linn., but abun- 
dant on both surfaces of the calyx of Christisonia neilgherrica 
Gardn.(52) and Aphyllon uniflorum Gray. (44) 

The mesophyll has a ground tissue of large, loosely packed, 
rounded to isodiametric, thin-walled cells with few to abundant 
starch grains in the inner half of the calyx. The va,scular 
bundles are located near its inner epidermis. 

The corolla is 2-lipped, tubular-campanulate (Plate 5, fig. 34), 
and its five lobes are more or less reniform, rotund or suborbic- 
ular, obscurely crenate, unequal or subequal, 0.25 to 1 centi- 
meter long, and about 0.5 to 1 centimeter, rarely 1.75 centi- 
meters, broad. It is twice, or even thrice, as long as the calyx, 

56,4 Jvliano: Bunga 427 

curved at the insertion of the stamens, 2.5 to 5 centimeters 
long. At its basal portion it is more or less constricted above 
the ovary. Externally the corolla is pale purple, deepening 
near the calyx, and white at its base. Internally it is deep 
purple in the region around the pistil and at its mouth. It is 
thinner and more delicate than the calyx, and usually dries off 

This corolla possesses an outer layer of epidermal cells (Plate 
3, fig. 23), which are rectangular, tangentially flattened, and 
with often papillate outer walls. The inner epidermal cells are 
much larger than the outer. From these epidermal cells of 
both inner and outer layers are developed numerous glandular 
trichomes (Plate 5, fig. 39) similar to those found on the calyx. 
Stomata are absent on the corolla of Aeginetia indica Linn., but 
present in Aphyllon nniflorum Gray. (44) 

The mesophyll consists of loosely arranged cells provided with 
starch grains, leaving between them large air spaces. Vascular 
bundles, which Me nearer its inner epidermis, traverse the 

The stamens are four in number (Plate 3, fig. 24), epipetalous 
and didymous (Plate 5, fig. 34), very similar to those found in 
ChristisoniaA^^) At maturity these are united at their anthers 
owing to the secretion of a thick mucilage by the glandular 
trichomes found on the filaments near their attachment to the 
connectives (Plate 3, fig. 26). The anthers are violaceous in 
color, and bilocular when young (Plate 3, fig. 24; Plate 5, fig. 
41), each locule being more or less crescent-shaped, owing to 
the projections of the broad, conical band of sterile tissue from 
the connective. At maturity of the anther, however, the wall 
partition between the two loculi breaks away, so that the anther 
becomes unilocular. Each of the two posterior stamens has a 
peculiar swollen prolongation, or spur, from the connective, 
which is deflexed and equal in length to the anther, arising near 
the attachment of the filament. This spur (Plate 3, fig. 25) is 
somewhat oblong in transverse section, and is filled internally 
with a parenchymatous ground tissue delimited by a distinct 
epidermal layer. The parenchymatous ground tissue possesses 
large intercellular spaces, and is traversed by a single large cen- 
tral vascular bundle that never reaches its apex. The filaments 
are white, rather slender, and more or less curved. Each of the 
filaments is somewhat polygonal in transverse section (Plate 3, 
fig. 26) and consists of large, rounded, parenchymatous ground 

428 The Philippine Journal of Science i93b 

tissue with large to small intercellular spaces and an abundance 
of starch grains. On the outer portion of the filament is a sin- 
gle layer of epidermal cells. On this epidermis are borne gland- 
ular hairs (Plate 5, fig. 39) similar to those formed on the 
calj^ and the corolla, and these hairs secrete mucilage cementing 
the anthers together long before anthesis. The glandular hairs 
are mostly at or near the attachment of the filament to the con- 
nective of the anther. Traversing the ground parenchyma of 
the filament is a single vascular bundle- 

The stigma is rather large (Plate 5, fig. 34), cordate-pileate, 
white to yellowish white, and 0.3 to 0.35 centimeter in diameter. 
It is covered with numerous hairlike processes, which are much 
elongated and somewhat pointed (Plate 4, fig. 32). Within 
these elongated cells, the cells are rather loose and contain an 
abundance of starch grains. On these expanded, elongated proc- 
esses of the stigma, the microspores germinate (Plate 4, fig. 32) . 

The style is somewhat cylindrical, inflexed, glabrous, white, 
hollowed in the center (Plate 3, fig. 27), and from 1.5 to 2 
centimeters long. It is covered by a distinct epidermis con- 
sisting of small, radially elongated, papillate cells, inclosing a 
ground parenchyma of loosely packed cells showing distinct air 
spaces, and provided with starch grains. Bordering on the 
central cavity are groups of elongated cells which are papillalike 
and swollen with mucilaginous secretions. These secretions are, 
perhaps, conducive to the growth of the male gametophyte. Two 
vascular bundles traverse the style throughout its length. 

The ovary is ovoid (Plate 5, fig. 34), conical near the stylar 
end, cordate at the base, pale yellow, glabrous, 1 centimeter long, 
and 0.7 to 1 centimeter in diameter. It is bicarpelled (Plate 3, 
fig. 24; Plate 5, fig. 46), and unilocular throughout. Borne on 
the wall of the ovary are two opposite, lamellate placentas, 
their free lamellae being irregular, deeply branched, and intensely 
plicate, becoming membranous at maturity of the fruit. These 
placentas extend into the ovarial cavity and lie through the whole 
length of the ovary. On the surface of the placentas are borne 
numerous anatropous megasporanges, which are extremely small. 
On the ground parenchymatous tissue of the ovary wall and 
the placentas is an abundant supply of starch grains. 

The unilocular ovary with four placentas is also reported in 
Christisonia bicolor Gardn.,(52) AphyUon uniflorum Gray, (44) 
and Epiphegus virginixma.iU) Of the four species of Aeginetia 
reported from the Netherland Indies by Bakhuizen van den 

56,4 Juliano: Bunga 429 

Brink, (3) Aeginetia indica Linn, has four placentas, the rest have 
only two. 

Organography of the flower. — The development of the floral 
organs is acropetal and they arise as follows: Calyx, corolla, 
stamens, and pistil. The earliest stage the writer was able 
to obtain was when the primordium of the pedicel to the flower 
has emerged from the axil of the scale leaf (Plate 2, fig. 14; 
Plate 5, fig. 35; Plate 8, fig. 89). It appears as a hemispherical 
mass of meristematic cells, which elongates first, and at its ter- 
minus the flower is developed. The first floral organ to emerge 
from this floral primordium is the calyx, and this is followed by 
the corolla (Plate 5, fig. 36) . Within the corolla, small mammil- 
late protrusions emerge as the primordia of the stamens (Plate 
5, fig. 37), and these are followed by the formation of the car- 
pels on the apex of the floral primordium. The calyx has a de- 
finite growth, so that the corolla has to break its way through 
it long before anthesis (Plate 5, fig. 34). 


The anther shows two distinct lobes in transverse section 
(Plate 3, fig. 24; Plate 5, fig. 41), and with a slight prolongation 
on the opposite side of the attachment of the filament. Under 
each lobe, one or more rows of cells (Plate 5, fig. 42) situated 
on the third layer below the epidermis, differentiate as the 
sporogenous tissue. These cells are easily distinguished from 
the surrounding cells by the density of their cytoplasm and by 
their heavily stained, large nuclei. This sporogenous tissue 
forms a solid crescent-shaped band on each lobe (Plate 5, fig. 
41) and is continuous through the whole length of the anther. 

Of the parietal tissue, which is two-layered (Plate 5, fig. 42) 
at first, the layer directly in contact with the sporogenous tissue 
functions as the tapetum. These tapetal cells elongate radially 
and become binucleate or even trinucleate at the time the micro- 
spore mother cells are in synapsis. The increase in the number 
of nuclei of the tapetal cells results from the divisions of their 

The outer parietal layer of cells (Plate 5, fig. 42) may in- 
crease in thickness by a few periclinal divisions during the 
development of the anther (Plate 5, fig. 43). As the micro- 
spores are formed, this parietal tissue is actually destroyed or 
consumed, except the hypodermal layer which forms the endo- 
thecium in the mature anther (Plate 5, fig. 49). 

430 The Philippine Journal of Science isas 

The sporogenous tissue by subsequent divisions forms a mass 
of microspore mother cells (Plate 5, fig. 43). As the micro- 
spore mother cell enters synapsis (Plate 5, fig. 44), its C3rtoplasm 
usually shrinks away from its original wall, and proceeds to 
divide successively to form the tetrahedral tetrads (Plate 5, 
fig. 45), at which time the original wall of the mother cell has 
already disappeared. The tetrads are at first embedded in a 
special gelatinous coat (Plate 5, fig. 45), and as this coating 
eventually dissolves (Plate 5, fig. 47), the microspores are 
liberated. The young microspores then round off, and acquire 
their own coats. Long before dehiscence the single nucleus of 
the individual microspore undergoes division so that it becomes 
binucleate (Plate 5, fig. 48). Each microspore then possesses 
a large rounded and distinct vegetative nucleus, and a small, 
lenticular, rather densely stained generative nucleus. Succes- 
sive division of microspore mother cells has also been reported 
in Epiphegus virginixma.iX^) Among other parasites in which 
binucleate microspores have been reported are: Viscum al- 
6wm,(4i,42) Rhopcdocnemis phalloides Jungh.,(32) Rafflesia 
patma B1.,(19) and Epiphegus virginianaAW However, in Rho- 
paiocnemis phalloides Jungh.(32) trinucleate microspores are 
sometimes present. 

Gates (20) has observed that the wall of the microspore mother 
cell in Lathraea usually remains in contact throughout the 
whole process of meiosis until its final dissolution. It does 
not round off and separate from its sister cells completely. 
Instead, a special microspore mother cell wall, often of great 
thickness, develops inside the original wall, and this special wall 
may lie in contact with or be clearly detachable from, the 
mother wall From this special wall wedge-shaped protrusions 
into the cytoplasm from four equidistant peripheral points be- 
come visible. After the formation of these wedges, active fur- 
rowing of the cytoplasm takes place, and the peripheral wedges 
are prolonged by passive deposition of wall materials and not 
by invagination. The prolongation of the peripheral wedges 
progresses towards the center. Simultaneously with this for- 
mation of wall from the special microspore mother wall, the 
original wall remains, and later dissolves, leaving the special 
"mother wall exposed, which in turn dissolves liberating the four 
microspores. The writer was not able to follow this very closely 
in Aeginetia, but the material on hand seems to indicate that 

56,4 JuHano: Bunga 431 

the same process of wall formation probably takes place in this 

Simultaneously with the destruction of the parietal tissue of 
each sporange, the cells separating the two adjacent micro- 
sporanges are also destroyed, so that the two microsporanges 
merge to form a single, large locule. At maturity of the anther 
only its epidermis and a single hypodermal cell layer, the en- 
dothecium, persist. The epidermis remains parenchymatous, 
while its endothecial cells become thick-walled and devoid of the 
rod-shaped thickenings that are often encountered in anthers of 
many angiosperms (Plate 5, fig. 49) . 


The megasporange appears as a small papilla formed by the 
activity of the cells of the placenta, a condition similar to that 
obtaining in Christisonia neilgherrica Gardn.(53) The epi- 
dermal cells of the placenta (Plate 6, fig. 50) contain dense 
cytoplasm and distinct nuclei, and are usually elongated vertical- 
ly. These epidermal cells show active anticlinal divisions at 
various points, and this activity affords a good means of al- 
lowing space for the growth of the young megasporanges. In 
this small papilla (Plate 6, fig. 50) the hypodermal cell of the 
placenta enlarges and divides periclinally (Plate 6, fig. 51). Its 
daughter cells enlarge and also divide periclinally (Plate 6, figs. 
52 to 54) . Division usually takes place first in the lower daugh- 
ter cell (Plate 6, fig. 52) and then in the upper cell (Plate 6, fig. 
53), or the upper daughter cell may not divide at all and three 
basal cells in a row may arise from the lower daughter cell of 
the juvenile megasporange (Plate 6, fig. 54) by the formation of 
two periclinal walls. However, the ultimate product is a unise- 
riate row of four or more cells (Plate 6, fig. 55), and these cells 
are covered with an epidermis, which is continuous with that 
of the placenta. The terminal cell (Plate 6, figs. 54 and 55) 
in the row usually functions as the megaspore mother cell, while 
the basal cells constitute the cells to the funiculus. 

As the megaspore mother cell elongates and enlarges, and by 
the time the young megasporange begins to show signs of un- 
equal growth, or even later, an epidermal cell of the megaspo- 
range divides periclinally (Plate 6, fig. 54). Division of this 
epidermal cell is followed by those behind it, such that the nucel- 
lus and the single integument become differentiated (Plate 6, 

432 The Philippine Journal of Science 1935 

figs. 56 and 57), the former remaining two-layered near the mi- 
cropyle, and only one-layered farther down the chalaza of the 
megasporange. Eventually, the integument becomes one-layered 
in the mature megasporange. 

As a result of a quicker growth of the cells on one side, the 
megasporange bends over towards the opposite side, its apex 
being at this time directed at an angle of about 45 "^ to the surface 
of the placenta (Plate 6, fig. 56). By further growth the apex 
of the nucellus becomes at length directed perpendicularly to 
the surface of the placenta, and the megasporange assumes its 
mature anatropous position (Plate 6, fig. 57). 

The development of a single archesporial cell in Orobanchaceae 
seems to be the rule ; an exception is found in Christisonia neiU 
gherrica Gardn.,(53) where occasionally two archesporial cells 
may differentiate, only one of which functions. In other para- 
sitic phanerogams so far studied, many exhibit a unicellular 
archesporium.(7, 17, 18, 19, 31) Of those parasitic plants develop- 
ing an archesporium consisting of more than one cell, may be 
mentioned the following: Loranthus sphaerocarpus,^^^) Viscum 
articulatum Burm.,(47) Helosis guyanensis,(i2) Balanophora 
elongata Bl.,(49) B. phalloides Jungh.,(32) Pedicularis foliosa 
Linn., Melampyrum silvaticum Linn., and Lathraea squamaria 
Linn. (39) 

As the anatropous condition of the megasporange is reached, 
the megaspore mother cell enters synapsis (Plate 6, fig. 56), and 
divides periclinally (Plate 6, fig. 57). By two successive divi- 
sions this megaspore mother cell gives rise to four megaspores 
in a row (Plate 6, fig. 58), of which the chalazal one becomes 
functional (Plate 6, fig. 59) . The formation of four megaspores 
has also been noted in OrobancheCJ) and Christisonia neilgherrica 
Gardn.,(53) of the Orobanchacese. 

In other phanerogamic parasites there is a great variation in 
the number of megaspores formed. For example, two mega- 
spores are formed in Dendrophthora oxycedri,{^^) Viscum 
albnm,(24t) and Balanophora elongata B1.;(17) three megaspores 
in Loranthus sphaerocarpus Bl.,(48) Arceuthobium oxycedri,(2S) 
and Pedicularis verticillata Linn.; (39) three to four megaspores 
in Balanophora globosa Jungh.,(3l) Rafflesia patma BL,(18) and 
Pedicularis verticillata Linnryi^Q) and four megaspores in Raf-- 
fiesta hasseltii T. and P., (19) Cytinus hypocistis Linn., (7) Brug- 
mansia zippelii,(W Euphrasia rostkviana Hayne, E. odontitis 
Linn., Alectorolophus hirsutus All., A. minor (Ehrh.) Wimm., 

s*»* Juliana: Bunga 433 

Pedicularis palustris Linn., P. caespitosa Sieb., P. recutita Linn., 
P. tuberosa Linn., P. /oZio^a Linn., Melampyrum silvaticum Linn., 
ilf. pratense Linn., To^^iia ai^prna Linn., and Lathraea squama- 
via Linn. (39) 

In some phanerogamic parasites exhibiting extreme degenera- 
tion of their plant organs there is a tendency to shorten the 
stages involved in the development of their megagametophytes. 
The formation of primary parietal tissue, for example, from the 
archesporium, which functions directly as the megaspore mother 
cell, is almost always absent in Loranthacese, (24, 48, 54) Balanopho- 
racese,(l7,3l,32) Rafilesiace8e(7,l8,i9) Scrophulariacea,(39) and 
Orobanchace8e;(53) in Arceuthobium oxycedn{2B) and Balano- 
phora elongata Bl.(49) the archesporium forms a primary parie- 
tal cell. In fact, in some of these parasitic plants the arche- 
sporium may even give rise to the embryo sac directly, as 
in LiZmm; (15) this is true also of Helosis guayanensis,0^) 
Rhopalocnemis phalloides Jungh.,(32) Balanophora elongata 
BL, (17, 49, 60) and Epiphegus virginianaAU) 

Of the four megaspores formed in Aeginetia indica Linn., the 
chalazal one becomes functional (Plate 6, fig. 59). Degenera- 
tion of the three micropylar megaspores does not follow a 
definite course. It may set in first at the middle megaspores 
(Plate 6, fig. 58), and then by the micropylar megaspore; or 
degeneration may start from the latter and proceed upward to 
the chalaza (Plate 6, fig. 59). At all events, only one of the 
four remains, and this is usually the uppermost, or chalazal 

The functional megaspore then enlarges and becomes much 
vacuolated. Its nucleus starts to divide quite early (Plate 6, 
fig. 60), and the daughter nuclei migrate to the opposite poles 
of the enlarged embryo sac, owing, perhaps, to the formation 
of a large central vacuole (Plate 6, fig. 61). Here the two 
daughter nuclei divide, and four nuclei (Plate 6, fig. 62) are 
formed. Another division of these four nuclei completes the 
normal octonudeate embryo sac often encountered among an- 
giosperms (Plate 6, fig. 63). In Orobanchaceae similar devel- 
opment of the embryo sac has been reported in Orobanche sp.(7) 
and Christisonia neilgherrica Gardn.,(53) and in Phelipaea coeru- 
lea{7) and Lathraea squamaria Linn., (39) of the Scrophulariacese. 

The mature embryo sac (Plate 6, fig. 63) is rather long and 
bottle-shaped, with its largest portion directed towards the mi- 

434 The Philippine Journal of Science 1935 

cropyle and the other end often reaching the single-layered 
integument at the chalazal region of the megasporange. At the 
micropylar region is located the egg apparatus, which consists 
of two synergids and a megagamete. The synergid is elongated, 
more or less pyriform, and possesses a large terminal and a small 
basal vacuole. It has a distinct nucleus embedded in a rather 
dense cytoplasm situated near its base. In Aeginetia, as well as 
in other phanerogamic parasites, the ''filiform apparatus" of the 
synergid is absent; in Santalum album(^^, 45) and Viscum 
ulbumM^) however, this structure is present. The synergid is 
ephemeral in character, and seldom persists long (Plate 6, fig. 64 ; 
Plate 7, fig. 67) in the embryo sac. It usually shows signs of 
degeneration at the time the primary endosperm nucleus is divid- 
ing (Plate 6, fig. 65) or even earlier (Plate 6, fig. 64; Plate 7, 
fig. 67). The megagamete is a little larger than either of the 
synergids, and lies between them. It has a larger basal vacuole 
(Plate 7, fig. 67) and a larger nucleus embedded in a dense 
peripheral cytoplasm at its apex. 

The polar nuclei (Plate 6, fig, 63) usually lie just above the 
egg apparatus and between the synergids or above the megaga- 
mete. They have large, distinct nucleoli and dense nucleoplasm. 
Soon these two polar nuclei fuse and a much larger fusion- 
nucleus results (Plate 6, fig. 64; Plate 7, fig. 67). This fusion- 
nucleus lies dormant for a long while above the egg apparatus, 
and it may be oblong, rounded, or oval, exhibiting prominent 
nucleoli and distinctly dense nucleoplasm. 

The antipodals (Plate 6, fig. 63) are situated at the extreme 
chalazal end of the sac in a socketlike projection, which nearly 
touches the integument. These antipodals are somewhat per- 
sistent and may be arranged in a row (Plate 6, fig. 63; Plate 
7, fig. 69) or the upper two may lie side by side below which 
is the third antipodal (Plate 7, figs. 67 and 70). In rare cases 
these antipodals may early show signs of degeneration (Plate 
6, fig. 63), but usually they persist long after endosperm forma- 
tion has been under way (Plate 7, figs. 69 and 70). Their cy- 
toplasm is not dense compared with that in the egg apparatus, and 
their nuclei are much vacuolated and poor in nuclear content. 
In Phelipaea coendea and Orobanche sp.(7) the antipodal cells 
are ephemeral in nature, while those found in Lathraea sqim- 
maria Linn. (7) are somewhat persistent, although not as per- 
sistent as those in Aeginetia indica Linn. 

56»4 Jvliano: Bunga 435 


Being ombrophytic in habit, this parasite flourishes in the 
thick cane plantation. Floral maturation usually takes place 
at the time the canes begin to flower or even earlier, at which 
time the sugar-cane plants have formed an impenetrable or im- 
passable cover around the buiiga plants in the fields. It will 
also be noted that the flowers of this root parasite are never held 
high from the surface of the ground. Under the conditions given 
above, the flowers of this parasite are seldom accessible to 
pollinating insects, except perhaps ants. The chance for insect 
pollination decreases as the distance from the edge of the cane 
field is increased, and vice versa. Insect pollination probably 
does take place, as it cannot be very well eliminated completely, 
but it may be considered as very insignificant, judging by 
the results of an examination of flowers at various stages of 

Long before anthesis, the filaments of the stamens are rather 
short, and the stigma is held very high. Simultaneously with 
the maturation of the flowers and usually just before anthesis, 
these filaments elongate further, and actually push their anthers 
clear to the stigma (Plate 5, fig. 34), brushing the microspores 
against it. It will also be noted that dehiscence of the micro- 
spores usually takes place long before anthesis, so that if the 
stigma is examined at the time the corolla has just opened, an 
abundance of germinating microspores will be seen on the stigma 
(Plate 4, fig. 32). Microspores that fail to lodge on the stigma 
may fall on the corolla tube, where they may germinate (Plate 
4, fig. 31) on account of the moisture present with the mucilagi- 
nous secretions within the corolla tube. 

Autogamy is, in the writer's opinion, the most potent factor 
in the pollination of this parasite; insects play only a minor 
role in the transfer of the microspores from the anthers to the 
stigma of the same or different flowers. In Orobanche(25) the 
flowers are homogamous, rarely protogynous, bee-pollinated, and 
sometimes devoid of nectar. Usually the stigmas are held high, 
but later the stamens grow to the same level with the stigmas 
and easily pollinate them, a condition which obtains in the buiiga 


Description. — The fruit of the bunga is a globose or ovoid 
capsule, which is light yellow when immature, turning ashy gray 

436 The Philippine Journal of Science isss 

or even jet-blaek at maturity. Its apiculate style is persistent 
and the whole fruit may be from 1.5 to 2 centimeters long and 
from 1 to 1.25 centimeters in diameter. When fully mature, it 
is brittle, and its pericarp easily breaks off, exposing its lamellate 
placentas on which are borne the yellowish, minute, rather pow- 
dery seeds. 

Development.-^Boon after fertilization the ovary enlarges and 
the corolla withers. The calyx and the corolla remain as a per- 
sistent envelope to the enlarging ovary within them. Of the 
pistil, the stigma is the first to show signs of darkening, and as 
the fruit enlarges this darkling proceeds downward to the style. 
The stigma usually falls off early, while the basal portion of the 
style together with the enlarging ovary form the major portion 
of the capsule, the calyx and corolla always remaining as a 
persistent envelope. At maturity of the capsule, the corolla and 
the calyx as well as the capsule become dry. The corolla and the 
calyx break off easily at a slight mechanical pressure from with- 
out, and this is true also of the pericarp, so the seeds are easily 
exposed. Dissemination of the seeds is made more perfect by 
the detachment of the scape from the mother parasite, and this 
is effected by the disintegration of the epidermal and cortical 
cells of the main axis of the scape at the region nearest the 
ground during drying. This breaking or separation is often has- 
tened by the attacks of termites, which are often abundant in 
cane fields. 

The ovary wall of the pistil is at first very thin, and consists 
mostly of thin parenchymatous ground tissue delimited by dis- 
tinct epidermal layers and provided with an abundance of starch 
grains. After fertilization and simultaneously with the matura- 
tion of the capsule, its pericarp becomes highly differentiated 
into two distinct regions. The cells of the outer half of the 
pericarp (Plate 5, fig. 40) remain parenchymatous. Later, the 
starch grains found in them disappear, and their walls become 
mucilaginous and ultimately disintegrate. On the other hand, 
the inner half of the pericarp possesses cells that become thick- 
walled and are highly lignified. This inner lignified half of the 
pericarp persists as the covering to the mature capsule. The 
mucilaginous character of the young capsule is partly due to the 
outer half of its pericarp and partly to the presence of the muci- 
lage secreted by the glandular hairs on the inner surface of the 
corolla. At maturity of the capsule the mucilage around it gets 

5«'4 JvUano: Bunga 437 

dry and the lignified inner tissue of the pericarp as well as the 
calyx and the corolla act as a protective wall to the fruit. 


Description.— The small, yellowish, albuminous seed, which is 
beautifully illustrated macroscopically by Kusano,(28) is irreg- 
ularly subglobose and is produced in abundance. Its testa con- 
sists of a layer of large, often yellowish cells with conspicuous 
reticulate thickenings. The cells of the testa are somewhat iso- 
diametric, much more developed at the chalazal end as well as 
on the sides of the seed, but poorly developed in the micropylar 
region (Plate 7, fig. 82). Inclosed by the single-layered testa is 
the endosperm, consisting of large isodiametric cells filled with 
starch grains and oil. If the seeds are immersed in concentrated 
chloral hydrate overnight and then pressed between a cover 
glass and a glass slide with a drop of Sudan III, numerous red 
fat globules will be seen oozing from the seeds. The cellulose 
inner and radial walls of the endosperm are thin, while their 
outer tangential walls are coated with a heavy, rather thin, 
cutinized layer, which is hardly soluble in concentrated sulphuric 
acid. Inclosed by this endosperm is a small, oblong or rounded 
embryo, consisting of one or often two rows of cells surrounded 
by a distinct epidermis. The cells of the embryo are also iso- 
diametric and parenchymatous, and in them are stored starch 
grains and oil. This embryo is simple in that it possesses no 
distinct plumule, radicle, or cotyledons. Its narrow end is di- 
rected towards the micropylar end of the seed. Kusano(29) also 
observed similar peculiarities of the seeds after placing them for 
some time in concentrated chloral hydrate and examining them 
in toto. 

Among numerous species of Orobanchaceae, Caspary(lO) 
has observed that the mature seeds of Orobanche cruenta Bert., 
O. procera Koch., O. pruinosa Lapeyr., 0. epithymum DC, 0. 
gain Duby, O. rubens Willd., 0. lucorum A. JBr., O. minor Sutt., 
O. amethystea Thuill., 0. cumana Wall., Phelipaea coerulea 
C. A. M., P. arenaria Walp., and P. ramosa C. A. M. are all egg- 
shaped, with one-layered testae from the chalazal end to the 
middle and three- to four-layered ones on the micropylar end. 
Orobanche has porous walls on the testa, the pores being nu- 
merous, small, and rounded in Orobanche cruenta, procera, prui- 
nosa, epithymum, galli, rubens, lucorum, minor, amethystea, and 
cumana. In Phelipaea the thickenings of the walls of the testa 

291458 3 

438 The Philippine Journal of Science 1935 

are reticulate, similar to those obtaining in Aeginetia indica 
Linn. Of the thirteen species examined by this investigator, 
the endosparm is egg-shaped and consists of from nineteen to 
twenty-one cells in longitudinal section, and these are filled with 
oil globules. At the micropylar region about three-eighths of 
its length is occupied by the embryo, which is ovate-elliptic and 
consists of thirteen to fifteen cells in longitudinal section, and 
is four to five cells wide; all the cells contain oil. The embryo 
is devoid of radicle, cotyledons, and other accessory organs. 

The seeds of the buiiga are produced in abundance, but many 
of them show the absence of embryos and are, therefore, sterile. 
The seeds also exhibit variations in size, and the sterile ones are 
usually smaller than the fertile plump seeds. In ovaries where 
numerous megasporanges are formed, it is to be expected that 
not all of these would be fertilized, and this fact should be 
kept in mind in considering the development of sterile seeds. 
Several other factors known to cause sterility in seeds have not 
been ascertained, but it is sufficient to know that a single seed 
is enough to cause trouble in a hill of canes. 

Endosperm. — The endosperm of this parasite is of the cellular 
type. Among the Orobanchacese a type of endosperm similar 
to the one reported here for the bunga has been described in 
Phelipaea coerulea and Orobanche sp.,(7) O. hederae,(^0) Chris- 
tisonia neUgherrica Gardn., and Ch. suhacavlis Giardn.,(53) and 
perhaps Orobanche cumana and O. ramasa. (40) Cicscutail^} 
and Rafflesia patma B1.(19) show a nuclear type of endosperm, a 
deviation from the great majority of the parasitic phanerogams 
so far studied. In Helosis guyanensisii^) the single polar nu- 
cleus gives rise to the pseudoendosperm. 

The primary endosperm nucleus very early divides (Plate 6, 
fig. 65) horizontally, and cytokinesis takes place between the two 
daughter nuclei (Plate 6, fig. 66). One of the daughter nuclei 
migrates into the small, cylindrical chalazal prolongation of the 
embryo sac, where it remains dormant (Plate 7, figs. 69 and 70). 
The lower daughter nucleus remains at the vicinity of the zygote, 
and this nucleus gives rise to the cellular endosperm found in 
the mature seed. 

The first division of the nucleus of the lower endosperm cell 
seems to be vertical, and by repeated horizontal walls (Plate 7, 
fig. 68) vertical rows of endosperm cells result (Plate 7, figs. 
69 and 70) • During the development of this cellular endosperm, 
the embryo sac enlarges considerably at its micropylar end 

66,4 Jvliano: Bunga 439 

where the endosperm remains one-layered (Plate 7, fig. 81), and 
widens rapidly upward where several layers of endosperm cells 
are produced (Plate 7, figs. 79 and 80) . The development of the 
cellular endosperm from the lower endosperm cell pushes the 
upper endosperm cell (Plate 7, figs. 69 and 70), clear to the per- 
sistent antipodal cells at the chalazal end of the seed. This up- 
per chalazal endosperm cell (Plate 7, fig. 70) may be homologous 
with the chalazal haustorium often encountered in other species 
of the Orobanchacese.(7) 

The micropylar cells of the cellular endosperm elongate con- 
siderably and form haustoriumlike processes on the sides of the 
zygote (Plate 7, fig. 70) . It is to be noted that although these 
micropylar endosperm cells have elongated longitudinally, they 
do not possess thick cytoplasm and distinctly amoeboid nuclei, 
so constant a feature in haustorial cells reported in many species 
of the Orobanchaceae. They seem to be degenerated haustorial 
cells (Plate 7, fig. 81), becoming nonfunctional in the species 
under discussion. Vertical divisions take place mostly among 
the endosperm cells (Plate 7, figs. 78 to 80) above these elongated, 
micropylar, haustorial cells, so that the endosperm becomes club- 
shaped in appearance, with its small, elongated, tapering lend 
directed towards the micropyle, and its enlarged portion towards 
the chalaza. Of the endosperm cells thus formed in the develop- 
ing seed, the peripheral ones are the largest, and are persistent 
(Plate 7, figs. 79 and 80) in the mature seed (Plate 7, fig. 82). 
Starch and oil are abundant in these endosperm cells. 

Bernard(7) has observed that the primary endosperm nucleus 
in Lathraea squamaria Linn, divides into two, a wall separating 
the primary endosperm cell into two equal parts, the chalazal 
cell nucleus dividing and the cell enlarging so that it becomes 
binucleate. The micropylar cell forms the cellular endosperm,, 
its micropylar cells becoming elongated and considerably enlarged 
with large nuclei. The enlarged binucleate chalazal endosperm 
cell elongates into the chalazal nucellus and funiculus, and pos- 
sesses dense cytoplasm around its nuclei. In Orobanche and 
Phelipaea the chalazal daughter endosperm cell elongates in the 
direction of the funiculus and forms a conducting element, typ- 
ical and homologous with that found in Lathraea. The superior 
endosperm cell becomes cellular. 

As the bunga seed matures, the embryo gradually absorbs the 
endosperm tissue so that nothing of this cellular endosperm 
remains except its outermost, peripheral layer (Plate 7, fig. 82). 

440 The Philippine Journal of Science 1935 

Embryo. — The zygote does not undergo division soon after 
fertilization (Plate 6, figs. 65 and 66; Plate 7, figs. 67 and 68), 
but remains dormant until after endosperm formation has long 
been underway (Plate 7, fig. 70). Soon after the cellular en- 
dosperm has been formed from the lower endosperm cell, the 
zygote elongates considerably, pushing itself between the endo- 
sperm cells (Plate 7, fig. 70). The zygote then seems to have 
undergone several transverse divisions (Plate 7, fig. 81), but 
the writer was not able to follow very closely the sequence of 
division of the zygote as the material on hand was very scant. 
It seems probable that the apical cell gives rise to the proembryo 
(Plate 7, fig. 71), while the basal cell forms the suspensor, which 
degenerates quite early (Plate 7, figs. 72 to 78). Division of the 
cells of the zygote is perhaps variable. However, a globose 
embryo devoid of cotyledons, radicle, and plumule, even of their 
rudiments, is developed in the mature seed (Plate 7, fig. 82). 

Nucellus. — The nucellus of the megasporange undergoes en- 
largement soon after fertilization. Its cells are isodiametric, 
vacuolated, and filled with an abundance of starch grains, and 
their nuclei are small. During later development of the seed 
the stored starch grains in the nucellus disappear and are ab- 
sorbed by the developing endosperm within. The enlarging 
endosperm then presses the nucellar cells against the massive 
testa, and this nucellus disappears in the mature seed. 

Coat or testa, — Before fertilization the single integument has 
two rows of cells near the micropylar end of the nucellus (Plate 
6, figs. 56 and 57) and only a single layer at the chalaza. Its 
outer layer of cells is continuous with the epidermis of the fu- 
niculus and placenta. As the seed approaches maturity, this 
coat becomes one-layered throughout. These cells of the coat at 
first form one of the tissues utilized for the storage of starch 
by the developing young seed, but after the endosperm is fully 
developed, this function is relegated to it, and starch grains dis- 
appear from the seed coat. 

Long before maturity of the seed the single-layered testa 
becomes lignified, and its cells develop the characteristic retic- 
ulate thickenings on their walls. This beautiful and thick retic- 
ulation of the cells of the single-layered coat of the seed of this 
parasite is carefully illustrated by Kusano.(28) These lignified 
and reticulated testa cells render the seed hard to section with 
the microtome. In other words, the mature seed is protected 
externally by the lignified and reticulated cells of the seed coat. 

^®'^ Jvliano: Bunga 441 

or testa, and internally by the cutinized layer surrounding the 
tangential walls of the endosperm. Neither the seed coat nor 
the endosperm is well developed at the mieropylar end of the 
seed, the former being sometimes absent or feebly developed, 
while the latter possesses no distinct cutinized layer on the outer 
walls of its cells. 


The anatomy and morphology of the vegetative and the re- 
productive organs of Aeginetia indica Linn., an important oro- 
banchaceous root parasite on the sugar-cane plant in the Philip- 
pines, are herein reported. 

An important factor to be considered as influencing its whole 
vegetative structure is its parasitism, and this finds expression 
in the retrogressive direction so as to suit the changed habit and 
conditions of the life of this parasite. Chief among its peculiar- 
ities, which may be taken as the outcome of its parasitic life, are 
(a) the branching and anastomosing root system; (6) the 
rhizomelike character of the roots, which are devoid of root 
hairs and root caps; (c) the formation of tubercles from which 
haustoria, young lateral roots, and scapes are produced— these 
tubercles serve as storehouses for the nutrition of the plant; (d) 
the modified anatomical structure of the cylinder of the root 
and scapes in which the xylem has become reduced and the 
phloem correspondingly developed; (e) the correlation of the 
reduction of its conducting system with the absence of foliar 
leaves replaced by scales; (/) the complete absence of chloropylU 
and (g) its great power of storing large amounts of starch 
in all its aerial as well as underground organs, which often pass 
out into the soil at the death of the plant. 

The anatomical structure of the root does not preclude the 
possibility of water absorption taking place from the soil. 

The root and the scape have endogenous origin, and the writer 
confirmed the findings of Kusano.(28) On ratoon canes second- 
ary roots may directly form scapes or reproductive organs later- 
ally or terminally, due perhaps to the stimulating action of the 
fire after the canes are harvested. 

The aerial organs, except the scale leaves, are devoid of naturar 
openings or stomata. The absence of natural covering to help 
prevent rapid loss of water from its organs is evident; this con- 
dition is responsible for its becoming an ombrophjrte, and its 
resulting inability to withstand extreme exposure to the sun. 

442 The Philippine Journal of Science 1935 

The development of its microspores is normal, and in all 
probability these microspores become separated by active fur- 
rowing of the cytoplasm through the formation of wedges derived 
from a special wall of the microspores similar to that obtaining 
in LathraeaA^O) The mature microspore is binucleate. 

The megasporange arises as a papilla from the placenta, and 
very early becomes anatropous. It is unitegumentary, and a 
single apical and terminal megaspore mother cell is differentiated 
on its nucellus. The mother cell develops normally into a con- 
ventional seven-celled megagametophyte. 

Pollination is largely autogamous, although insects may play 
a limited role in the transfer of its microspores from the anthers 
to the stigma of the same or different flowers. 

The fruit is a dry capsule, containing numerous minute, yel- 
lowish, albuminous seeds. Its endosperm is of the cellular type, 
exhibiting vestigial formation of haustorial cells. The seed coat 
consists of one-layered, lignified, reticulately sculptured cells, 
while its embryo is composed of a globose, few-celled structure, 
devoid of cotyledons, radicle, and plumule. Covering the endo- 
sperm is a highly cutinized layer. 

From the foregoing anatomical and morphological features 
of the bunga the following generalizations may be drawn : (a) 
Infestation of canes must be attributed largely to seeds left by 
previous buiiga crops, but other avenues of infection may be 
found; (5) the high drying power of its organs is due to the 
absence of any natural covering designed to limit the escape of 
moisture once it is exposed to air; (c) the darkening of the 
parasite on exposure to air is due to the presence of tannin ; (d) 
the seed is protected and is provided with an abundance of 
stored food (mainly starch and oil) in its endosperm and in its 
embryo, hence a very specialized form of attack has to be 
adopted to eradicate the parasite; (e) ratooning canes infested 
with the parasite is very favorable to the perennial growth of 
bunfa; and (/) the effort to eradicate it must be centered on 
preventing seedage and the germination of the seeds left by 
previous bunga crops. 


1. Agati, J. A., and J. P. Tan. The effect of atlacide on Aeginetia in- 

dica. Sugar News 12 (1931) 82-89. 

2. Agati, J. A., and J. P. Tan. Controlling the Aeginetia indica in cane 

fields. Sugar News 12 (1931) 852-854. 

5«r4 Jvliano: Bunga 443 

3. Bakhuizen van den Brink, R. C. Contribution a F^tude de la Acre 

des Indes Neerlandaises XXIII. Orobanchaceae in India Batavia 
Orientali crecentes. Bull. Jard. Bot. Buitenzorg III 13 (1933) 77- 

4. Barber, C. A. Studies on root parasitism. Mem. Dept. Agr., India 

1 (1906) No. 1, 1-30. 

5. Barber, C. A. Studies on root parasitism. The haustorium of San- 

talum album. Mem. Dept. Agr., India 1 (1907) No. 2, 1-58. 

6. Barber, C. A. Studies in root parasitism. The haustorium of Olax 

scandens. Mem. Dept. Agr., India 2 (1907) No. 4, 1-47. 

7. Bernard, Ch. Sur Tembryogenie des quelques plantes parasites. 

Journ. d. Botanique (Paris) 17 (1903) 23-32, 62-68, 117-133, 168- 

8. BissiNGER, G. H. Some of the problems now facing the Research 

Bureau of the Philippine Sugar Association. Sugar News 15 (1934) 

9. BOEWIG, H. The histology and development of Cassytha filiformis 

Linn. Contrib. Bot. Lab. Univ. Pennsylvania 2 (1901) 399-406. 

10. Caspary, R. Samen, Keimung, Species und Nahrpflanzen der Oroban- 

Chen. Flora 37 (1854) 577-588, 593-603. 

11. Chamberlain, C. J. Methods in Plant Histology. Chicago (1932) 


12. Chodat, R., and C. Bernard. Sur le sac embryonnaire de THelosis 

guyanensis. Journ. d. Botanique (Paris) 14 (1900) 72-79. 

13. COERT, J. H. Aeginetia spec, een wortelparasiet op het Suikerreit. 

Archief voor de Suikerindustrie in Nederlandsch-Indie 1924 (1924) 

14. COOKB, E., and A. F. Schively. Observations on the structure and 

development of Epiphegus virginiana. Contrib. Bot. Lab. Univ. 
Pennsylvania 2 (1901) 352-398. 

15. Coulter, J. M., and C. J. Chambmilain. Morphology of Angiosperms. 

New York (1903) 348. 

16. Elmer, A. E. D. Two hundred twenty-six new species-II. Leaflets 

Philip. Bot. 8 (1915) 2719-2883. 

17. Ernst, A. Embryobildung bei Balanophora. Flora N. F. 106 (1914) 


18. Ernst, A., and Ed. Schmid. Embryosackentwicklung und Befruch- 

tung bei Rafflesia patma Bl. Ber. d. Deutsch. bot. Gesell. 27 (1909) 

19. Ernst, A., and Ea Schmid. Uber Bltite und Frucht von Rafflesia. 

Morphologisch-biologische Beobachtungen und entwicklungsgeschicht- 
lich-zytologische Untersuchungen. Ann. d. Jard. bot. Buitenzorg 27 
(1913) 1-58. 

20. Gates, R. Pollen tetrad wall formation in Lathraea. La Cellule 35 

(1925) 49-59. 

21. GosECO, F. P. Preliminary report on Christisonia wightii Elmer, a 

new plant parasite on sugar cane. Sugar News 13 (1932) 73-74. 

22. HiNES, C. W. Diseases, insects and plant pests of the sugar cane in 

the Philippine Islands. PhOip. Agr. Rev. 11 (1918) 275-277. 

23. Johnson, T. Arceuthobium oxycedri. Ann. Bot. 2 (1888) 137-160. 

444 The Philippine Journal of Science 1935 

24. JOST, L. Zur Kenntniss der Bliitenentwicklung der Mistel. Bot. Ztg. 

46 (1888) 356-367, 372-887. 
25* Knuth, p. Handbook of Flower Pollination. Oxford 3 (1909) 644. 

Translated by J. R. Ainsworth. 

26. Koch, L. Ueber die Entwicklung des Samens der Orobanchen. Jahrb. 

wiss. Bot. 11 (1878) 218-261. 

27. KocH, L. Untersuchungen iiber die Entwicklung der Orobanchen. 

Ber. d. Deutsch. bot. Gesell. 1 (1883) 188-202. 

28. KUSANO, S. Notes on the Aeginetia indica Linn. Bot. Mag. Tokyo 

17 (1903) 71-95. (In Japanese; English resume on pp. 81-84.) 

29. KusANO, S. Further studies on Aeginetia indica. Bull. Coll. Agr., 

Tokyo Imp. Univ. 8 (1908) 59-78. 

30. Leclerc du Sablon. Recherches sur les organes d'absorption des plan- 

tes parasites (Rhinanthees et Santalac6es) . Ann. d. Sci. Nat. (Pa- 
ris) VII 6 (1887) 91-117. 

31. LOTSY, J. P. Balanophora globosa Jungh. Eine wenigstens ortlich 

verwittwete Pflanze. Ann. d. Jard. bot. Buitenzorg 16 (1899) 174- 

32. LOTSY, J. P. Rhopalocnemis phalloides Jungh. A morphological-sys- 

tematical study. Ann. d. Jard. bot. Buitenzorg 17 (1901) 73-101. 

33. Massee, G. On the structure and functions of the subterranean parts 

of Lathraea squamaria L. Journ. Bot. 24 (1888) 257-263. 

34. McWhortek, P. P. Concerning the sugar cane root parasite. Philip. 

Agriculturist 11 (1922) 89-90. 

35. Merrill, E. D. An Enumeration of Philippine Flowering Plants. Ma- 

nila 3 (1923) 628. 

36. RoxAS, M. L. The effect of "bunga" (Aeginetia indica) on sugar cane 

at Calamba Sugar Estate. Sugar News 8 (1927) 96-115. 

37. RoxAS, M. L. Aeginetia indica on sugar cane. Sugar News 12 

(1931) 89-91. 

38. SCHACHT, H. Die Bluthe und die Befruchtung von Santalum album. 

Jahrb. wiss. bot. 4 (1856-1866) 1-22. 

39. ScHMiD, E. Beitrage zur Entwicklungsgeschichte der Scrophularia- 

ceae. Beih. bot. Centralbl. 1 Abt. 20 (1906) 175-299. 

40. SCHNARF, K. Embryologie der Angiospermen. Berlin (1929) 673. 

41. ScHtiRHOFP, P. N. Die Befruchtung von Viscum album L. Ber. d. 

deutsch. bot. (Resell. 40 (1922) 314-316. 

42. ScHtJRHOFF, p. N. Die Zytologie der Bliitenpflanzen. Stuttgart 

(1926) 792. 

43. Scrobischewsky, W. Ueber den Ursprung des Fadenapparates bei 

Viscum album. Bot. Centralbl. 18 (1884) 156-157. 

44. Smith, A. C. The structure and parasitism of Aphyllon uniflorum 

Gray. Contrib. Bot. Lab. Univ. Pennsylvania 2 (1901) 111-121. 

45. Strassburger, E. Zu Santalum und Daphne. Ber. d. deutsch. bot. 

Gesell. 3 (1885) 105-113. 

46. Teodoro, N. G. Aeginetia indica in cane plantation. Philip. Agr. 

Rev. 18 (1925) 77-78. 

47. Trbub, M. Observations sur les Loranthac^es. Ann. d. Jard. bot. 

Buitenzorg 3 (1883) 1-12. 

56,4 Jvliano: Bunga 445 

48. Treub, M. Observations sur les Loranthacees. Ann. d. Jard. bot. 

Buitenzorg 2 (1885) 54-76. 

49. Trmjb, M, L'organe femelle et Tapogamie du Balanophora elongata 

Bl. Ann. d. Jard. bot. Buitenzorg 15 (1898) 1-39. 

50. Umiker, O. Entwicklungsgeschichtlich-cjrtologische Untersuchungen 

an Helosis guyanensis Rich. Diss. Zurich (1920). Cited by Schnarf. 

51. Wilson, L. L. W. Observations on Conopholis americana. Contrib. 

Bot. Lab. Univ. Pennsylvania 2 (1898) 3-19. 

52. WORSDELL, W. C. On the comparative anatomy of certain species of 

the genus Christisonia. Ann. Bot. 9 (1895) 103-136. 

53. WoRSDELL, W. C. On the development of the ovule of Christisonia, a 

genus of the Orobanchaceae. Journ. Linn. Soc. London 31 (1897) 

54. York, H. H. The origin and development of the embryo sac and 

embryo of Dendrophthora opuntioides and D. gracile. Bot. Gaz, 56 
(1913) 89-111, 200-216. 


Plate 1. Abginetia indica LiNNiEus 

Fig. 1. Transverse section of a young root; co, cortex; ep, epidermis; 
P ph, primary phloem; X 240. 

2. Transverse section of the central cylinder of an older root show- 
ing differentiation of the triarch primary bundles; ec, com- 
panion cell; pXf primary xylem; si t, sieve tube; co, cortex; 
X 156. 

8. Portion of a transverse section of a primary tubercle with a 
young primary root cut longitudinally, emerging from the for- 
mer; CO, cortex of the tubercle; X 156. 

4. Portion of a vascular bundle from a transverse section of a 

mature root (taken from fig. 7) ; ca c, cambial cell; px, primary 
xylem; scl r, sclerenchymatous ring; si t, sieve tube; tr, tra- 
cheid; X 156. 

5. Diagram of a section of a primary tubercle showing the cortex 

(co)^ central cylinder or stele (sie), primary root primordia 
(pr), and vascular bundles {vh)\ X 18. 

6. Diagram to show relationship between the host root (hr) and the 

parasite; co, cortex of tubercle; sie, stele of tubercle; h, haus- 
torium; x 18. 

7. Diagram of a transverse section of a mature root; co, cortex; 

epy epidermis; px, primary xylem; scl r, sclerenchymatous ring; 
s phy secondary phloem; sx, secondary xylem; X 18. 

8. Embryogenic cells of a primary root primordium (shaded cells) ; 

CO, cortex of tubercle; X 156. 

9. Portion of a transverse section of the scale leaf; I ep, lower epider- 

mis; u ep, upper epidermis; vh, vascular bundle; X 74. 

Plate 2. Aeginetia indica Linnaeus 

Fig. 10. Transverse section of a secondary root showing early formation 
of secondary tubercle; co, cortex; ep, epidermis; p ph, primary 
phloem; px, primary xylem; X 38. 

11. Portion of a transverse section of a young primary tubercle show- 

ing vascular bundles cut transversely; X 156. 

12. Longitudinal section of a haustorium (taken from Plate 1, fig. 

6); CO, cortex of tubercle; en, endodermis of host root; tr, 
tracheid to the haustorium; xy, xylem vessel of the host root; 
X 156. 

13. Showing epidermal cells of the haustorium in contact with the 

cortical cells of the host root; ep, epidermal cells of the hausto- 
rium; hr, host root; X 156. 


448 The Philippine Journal of Science 1935 

Fig. 14. Diagram of a transverse section of a tubercle showing a scape 
arising from it; co, cortex; ep, epidermis; /p, floral primordium; 
al, scale leaves; tub, tubercle; X 18. 

15. Diagram of a transverse section of the host root with a tubercle 

attached to it; co, cortex; tub, tubercle; h, haustorium; hr, 
host root; ste, stele of the host root; X 18. 

16. Portion of a transverse section of the main axis of the scape 

showing a single vascular bundle; co, cortex; bs, bundle sheath; 
ep, epidermis; ph, phloem; pi, pith; tr, tracheids; X 156. 

Plate 3. Aihsinbtia indica Linnaeus 

Fig. 17. Diagram of a portion of a transverse section of the main axis 
of the scape far from the tubercle showing positions of the 
vascular bundles; co, cortex; ep, epidermis; 6s, bundle sheath; 
vb, vascular bundle; X 18. 

18. Portion of a single vascular bundle taken at A from fig. 17; si t, 

sieve tube; tr, tracheid; X 156. 

19. Portion of a single vascular bundle taken at B from fig. 17; si 

t, sieve tube; tr, tracheid; X 156. 

20. A diagram of a transverse section of the scape near its apex 

to show positions of the scale leaves and the pedicels; m ax, 
main axis; ped, pedicel; si, scale leaves; X 18. 

21. Diagram of a transverse section of a young pedicel showing 

positions of procambial strands (pa) ; co, cortex; ep, epider- 
mis; X 18. 

22. Portion of a transverse section of the calyx; i ep, inner epider- 

mis; vb, vascular bundle; X 156. 

23. Portion of a transverse section of the corolla; i ep, inner epi- 

dermis; o ep, outer epidermis; vb, vascular bundle; X 156. 

24. Diagram of a transverse section of a young flower showing the 

relative positions of its parts; ca, calyx; cor, corolla; I, locule; 
ov, ovary; p, placenta; st, stamen; x 18. 

25. Diagram of a transverse section of the spur of the stamen ; ep, epi- 

dermis; vb, vascular bundle; X 18. 

26. Diagram of a transverse section of a filament; ep, epidermis; gh, 

glandular hairs; vb, vascular bundle; X 18. 

27. Diagram of a transverse section of the style; ep, epidermis; mc, 

mucilage cells; ho, hollow; vb, vascular bundle; X 18. 

Plate 4. Abgineha indica Linnjeus 

Fig. 28. Portion of a transverse section of an old root; co, cortex; ph, 
phloem; px, primary xylem; ax, secondary xylem; scl r, scle- 
renchymatous ring; X 93. 

29. A portion of a transverse section of a mature pedicel; co, cortex; 

ph, phloem; x, xylem; pi, pith; X 93. 

30. Longitudinal section of a root showing its apex; X 93. 

31. Showing germinating microspores on the inner surface of the 

corolla tube; X 93. 

32. Portion of a longitudinal section of the stigma showing micro- 

spores germinating on its papillate cells; X 93. 

5«,4 Jvliano: Bunga' 449 

Fig. 83. A portion of a transverse section of the main axis of the scape 
near its attachment to the tubercle; co, cortex; vb, vascular 
bundle; pi, pith; X 62. 

Plate 5. Aegin^ttia indica Linn-sus 

Fig. 34. A sketch of a mature flower with half of its calyx and corolla 
removed so as to expose its other organs; ca, calyx; cor t, corol- 
la tube; ov, ovary; ped, pedicel; st, stamen; stiy stigma; s^.y, 
style; X 0.66. 
85. A diagram of a longitudinal section of a young scape showing 
scale leaves (si) at the axils of which arise the floral primor- 
dium (fp); X 18, 

36. Longitudinal section of a young floral primordium showing calyx 

(ca) and beginnings of the petals (pe) ; x 18. 

37. An older flower showing calyx (ca), petals (pe), and stamens al' 

ready differentiated (st) ; X 18. 
88. Showing young hairs from the inner epidermal cells of the calyx; 
X 360. 

39. Mature glandular hairs from the inner epidermis of the calyx; 

X 360. 

40. Portion of transverse section of the ovary wall, or pericarp, of a 

nearly mature capsule showing its lignified inner half and a 
disintegrating, mucilaginous outer half; x 66. 

41. A diagram of a transverse section of an anterior anther; ap t, 

sporogenous tissue; vb, vascular bundle; X 38. 

42. Showing the sporogenous tissue (sp t) and tapetum (t) with 

the parietal tissue (pt) (taken from fig. 41); ep, epidermis; 
X 360. 

43. Portion of a longitudinal section of an older anther showing the 

mass of microspore mother cells (mi mc), tapetum (t), and thin 
parietal tissue (pt) ; ep, epidermis; X 360. 

44. Microspore mother cell in open spireme; X 660. 

45. Tetrads still in a special mucilaginous coat; microspore mother 

cell wall already gone; X 660. 

46. Transverse section of the ovary showing the four placentas (p) ; 

I, locule; X 18. 

47. Tetrads without special mucilaginous coat; X 660. 

48. Microspore just before dehiscence showing a large, rounded pollen 

tube nucleus (ptn), or vegetative nucleus, and a lenticular gen- 
erative nucleus (gn) ; X 660. 

49. Portion of a longitudinal section of a mature anther showing 

epidermis (ep), endothecial layer (endt), and remains of the 
parietal tissue (pt) ; note the size of the microspores (mi) ; 
X 240. 

Plate 6. Aeginetia indica Linn^us 

Fig. 50. Showing enlargement of a hypodermal cell of the placenta and 
anticlinical division of its epidermal cell (ep) ; X 360. 
51. First anticlinal division of a hypodermal cell of the placenta; 
X 360. 

450 The Philippine Journal of Science 1935 

Fig. 52. Periclinal division of the lower daughter nucleus of the placental 
hypodermal cell; X 360. 

53. Upper daughter cell dividing; X 360. 

54. Young megasporange with an elongated, terminal megaspore 

mother cell (mmc) ; three lower cells arising from the lower 
daughter cell of the hypodermal initial; note the beginnings of 
the integument (in) ; X 360. 

55. An older megasporange showing the megaspore mother cell; note 

unilateral growth; X 360. 

56. A megasporange showing its terminal megaspore mother cell in 

open spireme; integument (in) fully differentiated as well as 
the nucellus; mic, micropyle; /, funiculus; X 360. 

57. A still older megasporange with megaspore mother cell in meta- 

phase stage (heterotypic); in, integument; /, funiculus; ti, nu- 
cellus; X 360. 

58. Tetrads; two middle megaspores showing degeneration, chalazal 

one enlarging; X 360. 

59. Showing chalazal megaspore enlarged; three micropylar mega- 

spores degenerated; X 360. 

60. First division of the nucleus of the functional megaspore; X 660. 

61. Binucleate embryo sac showing polarity; X 660. 

62. Quadrinucleate embryo sac; X 660» 

63. Mature embryo sac showing contents; antipodals (an) showing 

early degeneration; polar nuclei (pn) still unfused near the 
egg apparatus; me, megagamete; sy, synergid; X 660. 

64. Showing early degeneration of the synergid (sy); fn, fusion nu- 

cleus; me, megagamete; X 660. 

65. Showing the first division of the primary endosperm nucleus 

and degeneration of the synergid (ay); zygote {zy) dormant; 
X 660. 
^Q, Daughter cells from the first division of the primary endosperm 
cell; zygote (zy) still undivided and synergid (sy) degenerat- 
ing; I end c, lower endosperm cell; u end c, upper endosperm 
cell; X 660. 

Plate 7. Aeginbtia indica LiNNiBus 

Fig. 67. Embryo sac showing degenerated synergid (sy) and antipodals 
(an) ; fn, fusion nucleus; me, megagamete; X 660. 

68. Showing the second division of the daughter endosperm nuclei; 

wall not cut across; zy, zygote; sy, synergid; X 660. 

69. Chalazal portion of the endosperm showing antipodals (an) and 

the upper endosperm cell (u end e) ; X 660. 

70. Embryo sac with cellular endosperm, micropylar cells of which 

have enlarged and elongated; upper endosperm cell (u end c) 
remaining inactive and much elongated; antipodal nuclei (an) 
persisting and zygote (zy) elongated considerably; x 660. 
Figs. 71 to 78. Proembryos with degenerated suspensors (sua); X 660. 
79 and 80. Transverse sections of the endosperm (end) and embryo 
(em); X 660. 

56,4 JuLijam): Bunga 451 

Fig. 81. Longitudinal section of an older endosperm (end) showing the 
enlarged micropylar cells; note the embryo (em) with dark- 
ened nuclei; X 360. 
82. Longitudinal section of a nearly mature seed (seed coat not 
included); end, endosperm; cl, cutinized layer; em, embryo; X 

Plate 8. Aeginetia indica Linnaeus 

Fig. 83. Portion of a transverse section of a young root showing early be- 
ginning of a branch root (cells with darkened nuclei) ; co, 
cortex; pi, pith; p ph, primary phloem; X 240. 

84. Portion of a transverse section of an older root showing the 

primordium of the branch root within the cortical tissue , of 
the mother root; X 240. 

85. Transverse section of a secondary root showing more compact 

and active cortical cells on its lower side; note the spongy 
character of the cortical cells on the opposite side; ep, epi- 
dermis; CO, cortex; p ph, primary phloem; pi, pith; x 240. 

86. Further differentiation of the lower cortical cells of a secondary 

root in the formation of the secondary tubercle; note the vas- 
cular bundles (vb) ; X 240. 

87. Portion of a section of a primary tubercle showing primordium of 

a scape still within the thin cortical tissue (co) ; si, scale leaf 
rudiment; X 75. 

88. An older scape emerging from the primary tubercle with young 

lateral scale leaf (si); vb, vascular bundle; X 75. 

89. Diagram of a section of a primary tubercle showing young 

scape emerging from its lateral side; axillary pedicel (ped) 
already differentiated; m ax, main axis to the scape; si, scale 
leaves; X 18. 

Plate 9. Aeginetia indica Linnaeus 

Fia 90. Ratoon canes (POJ 2878) dug April 28, 1934, with fresh par- 
asite producing young scapes. Reduced to about one-half its 
natural size. 
91. A, a portion of a root of the parasite developing scapes (sc) 
at its side and injured apex without forming tubercles, X 4. 
B, tubercle (tub) forming a scape (sc); (br) branch roots, 
X 2. 

Plate 10. Aeginetia indica LiNN-ffius 

Fig. 92. The parasite in bloom on ratoon canes (POJ 2878) June 14, 
1934, about two and one-half months after the canes had been 
harvested. Reduced to about one-half its natural size. 
93. The growth of the parasite (pa) one month after the seedling 
(hs) cane (POJ 2878) had been inoculated with fresh viable 
seeds. The specimen was preserved in 10 per cent formalin 
before it was photographed. Reduced to about one-half its 
natural size. 

.lriJAN«i: lii'Nr.A.I 

I 1*1111. II*. .Im'RW S<i,, 7^(1, N«>. !. 



{ l^ifiLiP, -JorRN. 8ri,, .ifi, Nil. -1. 


Jli.iANo: |irr??;A.| 

I Philip. J«»i rx', Sci., r»il. No. I 


IFtltLlP. JOLTRN. SCl.. 


fC,:^. l-'^'^''^M^":'i^^^} 

,. X. ^^^5^ . 



' % 


C€j ; 

^ \--4Y 


* y r '■■■ 


'C ^■^.«!* t"V., 

Jl'I.IANCK- Br'\TL4 f 

I PiiiMP. liiviiw Sri,, m, N«. 4. 


jMr.lANo: ffltNClA.I' 

f\ J«I UK. Sf'!., r.ti, HiK k 


JuLlANd: lirNTjA.I 

I FiiiLiF. JiH'KN. Sri., 5r», N<i. 4. 


fH I lAKO r }!i"V(;a.; 

I J.H'R?^. S<'L, uts N«i 


1 I \s^->r'^^* 

\ ■ 

pp/l - 

>w ' 



'V:^^V-"^^i\Wv :.>■OVr 



[PlULIP, .lOUKN. SCI., 5l», No, t. 





By Oakes Ames 

Professor of Botany in Harvard University 


Eduardo Quisumbing 

Curator, Philippine Natiorml Herbarium, Bureau of Scienee, Manila 


The present paper is essentially similar to its predecessors.^ It 
consists of descriptions of one new species* and four new va- 
rieties. Four previously described species, Dendrobium cerinum, 
Bulbophyllum Lobbii, Renanthera elongata, and Trichoglottis 
Guibertii, axe for the first time credited to the Philippine Is- 
lands with certainty. Dendrobium anosmum var. Dearei and 
Dendrobium anosmum var. Huttonii, which were known only 
from their original descriptions, were recently rediscovered. 
Dendrobium Schuetzei, which is known only from Agusan and 
Surigao Provinces, Mindanao, is for the first time illustrated in 
color and redescribed. All the descriptions in the text have been 
prepared from living specimens, and all the illustrations were 
made by Mr. Pedro Ramos, draftsman of the National Museum 
Division, Bureau of Science. All the types of the new species 
and varieties have been deposited in the Philippine National 
Herbarium, Bureau of Science, formerly called Herbarium of 
the Bureau of Science, with isotypes in the herbarium of the 
senior author. Available isotypes will be distributed to the 
herbarium of the New York Botanical Garden and to other 

Genus DENDROBIUM Swartz 

DENDROBIUM ANOSMUM Lindl. var. DEAREI (Rolfe) Ames and Qnls. eomb. nor. 
Plate 1^ fiff. 1; Plate 5. 
Dendrobium superbum Reichb. f. var. Dearei ROLFB in Lindenia 6 
(1891) 52, sub t. 264; Rolfe in Orch. Rev. 14 (1906) 177, fig. 22; 
Ames, Orch. 2 (1908) 187. 

* Philip. Journ. Sci. 44 (1931) S69-S83, 16 pis.; 47 (1932) 197-220, 
29 pis.; 49 (1932) 483-504, 28 pis.; 52 (1933) 443-473, 17 pis. 

291438 4 453 

454 The Philippine Journal of Science 1935 

? Dendrohium macranthum Hook. var. album Naves, Novis. App, 
(1882) 233. 

La variete Dearei a les fleurs du blanc le plus pur, avec les segments 
tres pointus. 

In habit similar to the species, but the stems are shorter and 
more slender. The flowers are pure white, except the naphtha- 
lene yellow throat of the lip, slightly fragrant, 8.5 to 9.5 cm 
across. The sepals lanceolate, 5.2 to 5.3 cm long, 1.5 to 1.6 cm 
wide, the dorsal acute, the laterals acuminate and forming a 
mentum or short spur which is pale greenish. Petals elliptic, 
acute, about 5 cm long, 3.2 cm wide. 

Luzon, Manila, Mrs. Remedies C. Gonzales's gardens, Phil. 
Nat Herb, 102 Quisumbing, March 9, 1934. This is the second 
time that this variety has come to the attention of the junior 
author. The first instance was a plant belonging to Mrs. Ste- 
wart, collected by her near Baguio in March, 1926. The second 
plant, the subject of this description, was originally collected 
from the mountains of Rizal Province and was found mixed 
with typical Dendrobium anosmum. 

This is one of the purest white dendrobes known, and is 
endemic to the Philippines. 

DENDROBIUM ANOSMUM Lindl. var. HUTTONII (Reichb. f.) Ames and Qais. comb, 
nov, Plate 1, fiip. 3. 

Dendrobium superbum Reichb. f. var. Huttonii Reichb. p. in Gard. 
Chron. (1869) 1206; Rolfe in Lindenia 6 (1891) 52, sub. t. 26U; 
Orch. Rev. 29 (1921) 116; Williams, Orch. Grow. Man. ed. 7 
(1894) 364; Ames, Orch. 2 (1908) 187. 

Perigonio candido, labelli disco ac ungue purpureis. 

A very striking novelty. The flowers are clear white. The disc of 
the lip bears two beautiful purplish blotches, and its base is of the same 
colour. For this beautiful thing we have to thank Messrs. Veitch, who 
obtained it through their late excellent collector, Mr. Hutton, from the 
Malayan Archipelago. 

In habit and general features similar to the species. The 
stems are comparatively shorter and more slender. The flowers 
are beautiful, showy, slightly fragrant, and about 8.5 cm across. 
They resemble those of the former variety in being pure white 
throughout but have the throat of the labellum dark purple. 

Luzon, Manila, Mrs. Remedios C. Gonzales's gardens, Phil. 
Nat. Herb. lOS Quisumbing, March 10, 1934. The plant was 
originally collected by orchid peddlers among the typical plants 
of Dendrobium anosmum, from the mountains of Bizal Prov- 
ince, Luzon. 

56» 4 Ames and Quisumbing: Philippine Orchids, V 455 

This, like the former variety, is very rare and beautiful, hav- 
ing been seen by the junior author in Manila gardens but once 
during his many years of study of orchids. It is characterized 
by having white flowers with the throat of the labellum dark 
purple. This particular variety has also been reported from 
the Malay Archipelago. 

DENDROBIUM CERINUlf ReichK f. Plate 1, fiera. 4 and 5; Plate 3, figs. 1 to 10; Plate 
9, fig. 2. 

Dendrobium cerinum Reichb. p. in Gard. Chron. II 12 (1879) 554. 

Dendrohium cerinum, n. sp. — Caule teretiusculo demum multum sulcato 
calamum aquilinum crasso; racemis brevibus paucifloris densifloris; sepalo 
imparl ovato oblongove acutiusculo; sepalis lateralibus triangulo semi- 
ovatis obtuse acutis, in mentum teretiusculum apice abrupte acutum ex- 
tensis; tepalis ovatis obtuse acutis; labelli ungue cum cornu retrorso in 
medio, lamina subrotunda laevi, antice minute denticulata, columna cla- 
vata — ^Flores cerini, nitidi, ochroleuci, brunneo lavatis. Lineae fuscae ra- 
diantes in bast laminae labelli. — ^Ex aff . Dendrobii sanguinolenti ; labello 
tamen ac mentum multum recedens. Ex archipelago Sondaico viv. misit 
el. Burbidge ad dominos Veitch. 

This is very near the well-known buff-coloured variety of Dendrobium 
sanguinolentum. The stem is half as thick as one's little finger, and much 
furrowed when old. It bears dense racemes of from four to six flowers, 
whose chin is thinner and abruptly acute, not thick and retuse as in the 
compared species. Ovaries and pedicels light rose. Sepals and petals 
light yellowish-ochre coloured, shaded with brown. The central rib of 
the mentum is light purple. The lip is just alarming. It mimics that of 
Dendrobium sanguinolentum, having a strong retrorse tooth on its claw. 
The blade, however, shows the best differences; it is oblong, not three-lobed, 
and its border, in lieu of being totally entire, shows numerous minute 
teeth on its anterior edge. The colour is light ochre with radiating brown 
lines at the transition of the blade into the claw. The whole flower is 
of very firm texture, and shining as if made from wax. The lip has 
sometimes a certain tendency to become three-lobed. I have to thank for 
materials Messrs. Veitch, who tell me that the plant was collected in the 
Malayan Archipelago by Mr. Burbidge. It flowered in July and Sep- 
tember. H. G. Reichenbach f. 

Stems terete, aggregated, subpendulous, fusiform, 4.5 to 17 
cm long, 5.5 to 10 mm in diameter at the widest portion, angled 
and sulcate when dry, the nodes 1 to 2.5 cm distant. Leaves 
distichous, elliptic-lanceolate or oblong-lanceolate, 6 to 11 cm 
long, 1.8 to 3.2 cm wide, nervose, membranaceous, pale green, 
narrowed to the acute apex. Leaf sheaths green, membrana- 
ceous, deciduous on the lower part of the plant. Racemes short, 
pendulous, laxly few-flowered, up to 4.5 cm long, 2- to 3-flowered. 
Flowers 3.5 to 3.8 cm long, 3.5 to 3.8 cm across, odorless, with 
slightly inflexed spur. Bracts pellucid, minute, membranaceous, 

456 The Philippine Journal of Science i985 

1.5 to 2 mm long. Pedicellate ovary 2.3 to 2.5 cm long, slender. 
Lateral sepals very obliquely oblong-lanceolate, acute, 1.4 to 1.6 
cm long, 7-nerved, forming with the column foot a mentum or 
spur which is elongate, slightly curved, obtuse, 1.9 to 2 cm long. 
Dorsal sepal lanceolate-ovate, subacute, 1.4 to 1.6 cm long, 7 to 
8 mm wide, 7-nerved. Petals spreading, oblong-^elliptic, rounded 
and minutely apiculate at the apex, 14.5 to 15.5 mm long, 7.5 
to 8 mm wide, 5-nerved. Labellum simple, elongate, about 3 
cm long, attached to the column-foot; claw linear-oblong, sharply 
sigmoid, with a retrorse tooth in the middle ; lamina round-ovate, 
about 1.8 cm long, 1.7 cm wide at the broadest portion (when 
expanded), broadly rounded or slightly retuse at the apex, cren- 
ulate-denticulate on the margin; disc provided with a single 
broad rather obscure central fleshy band which is glabrous and 
extends from the column-foot nearly to the apex of the labellum. 
Column very short and stout, tridentate. Anther subquadrate- 
ovoid, about 3 mm long. 3.5 mm wide. 

Luzon, Manila, Bureau of Science orchid house, PhiL Nat. 
Herb. 104 Quisumbing, February 28, 1933. The living plants 
were originally gathered by Novaliches orchid peddlers from 
the mountains of Rizal Province, Luzon, back of the town of 
Antipolo. The plants are now being cultivated in the Bureau 
of Science orchid house and in the gardens of Mrs. Remedios C. 
Gonzales. The description was based on Bureau of Science ma- 
terial which flowered in Manila February 28, 1933. The flowers 
are fairly large and odorless, and remain fresh for three to 
four days. Pedicellate ovary apple green ; sepals straw yellow ; 
petals and labellum chalcedony yellow; spur straw yellow with 
shades of ecru-olive and some very pale purple, light lumiere 
green at the tip. 

In the herbarium of the senior author is a colored sketch of 
this species from the Reichenbachian collection in Vienna. 
Thanks to this authentic record, one more obscure species is 
now shown to be represented by a recent definite collection. 

Apparently the original collection was destitute of leaves. It 
was described (and shown) as having light rose or reddish 
pedicellate ovaries whereas the recent collection has the pedi- 
cellate ovary apple green. 

A species closely allied to Dendrobium Guerreroi Ames and 
Quis. but differing from it in its fusiform, short, angled stems, 
in its few-flowered racemes, in the color of the flowers, in the 

56. 4 Ames and Quisumbing: Philippine Orchids, V 457 

shape of the spur, and in its denticulate-margined labellum 
with ovate-rounded lamina. 

DENDROBIUH PHILLIPSII sp. nov. Plate 1, figs. 6 and 7; Plate 3, figs. 11 to 19; Plate 
6, fisr. 1. 

Aff. D. Fairchildae. Caules penduli, aggregati, teretes, te- 
nuiter-fusiformes, 45 ad 60 cm longi, 5 ad 7 mm in diametro. 
Folia lanceolata, acuta, papyracea, decidua, 11.5 ad 12.5 cm longa, 
1.5 ad 2 cm lata. Racemi brevissimi, pauciflori; bracteae lan- 
ceolatae, acutae, 5 ad 6.5 mm longae. Sepala lateralia obliquis- 
sime triangular i-ovata, acuta, 1.6 ad 1.7 cm longa, columnae 
pedem secundum 1.2 ad 1.3 cm lata; mentum valde elongatum, 
obtusum, ad apicem haud inflexum 2.7 ad 3 cm longum. Sepalum 
dorsale ovato-lanceolatum, subacutum vel obtusum, 1.8 ad 1.9 cm 
longum, 9 ad 10 mm latum. Petala late oblonga vel oblongo- 
elliptica, apice rotundata, 1.7 ad 1.8 cm longa, circiter 8 mm 
lata. Labellum simplex, ovato-oblanceolatum cum parte infe- 
riore oblonga sensim dilatata, antice ovato-triangulare, acutum, 
circiter 3.5 cm longum et 1.5 cm latum. Gynostemium brevis- 
simum, crassum, apice tridentatum cum stelidiis lateralibus trian- 
gulari-ovatis leviter recurvatis et dente medio tenui, in pedem 
elongatum extensum. 

Stems pendulous, crowded, terete, elongate, fusiform, 45 to 60 
cm long, 5 to 7 mm in diameter, reddish-brown, smooth; inter- 
nodes 2.5 to 4.5 cm long. Leaves lanceolate, acute, 11.5 to 12.5 
cm long, 1.5 to 2 cm wide, papery, deciduous. Racemes short, 
few-flowered, 3 to 4 cm long; bracts lanceolate, acute, greenish, 
5 to 6.5 mm long, 3 to 3.5 mm wide when expanded. Flowers 
about 1.8 cm across. Pedicellate ovary about 3 cm long. Lat- 
eral sepals very obliquely triangular-ovate, acute, 1.6 to 1.7 
cm long, 1.2 to 1.3 cm wide along the column-foot; the mentum or 
spur stout, elongate, obtuse, straight, 2.7 to 3 cm long, 1.2 to 1.3 
cm wide near its base. Dorsal sepal broadly ovate-lanceolate, 
subacute or obtuse, 1.8 to 1.9 cm long, 9 to 10 mm wide. Pe- 
tals broadly oblong or oblong-elliptic, rounded at the apex, 1.7 
to 1.8 cm long, about 8 mm wide. Labellum simple, ovate-oblan- 
ceolate with the lower part gradually narrowed and adorned with 
a low transverse lacerate ridge, about 3.5 cm long, triangular- 
ovate near the apex, acute, about 1.5 cm wide. Gynostemium 
very short, broad, with the lateral stelidia triangular-ovate and 
slightly recurved, and middle tooth slender. Foot elongate. 

458 The Philippine Journal of Science 1935 

Luzon, Manila, Bureau of Science orchid house, Bur. Sci. 85618 
Quisumbing, February 14, 1933. The above description was 
based upon a living flowering specimen sent to the junior author 
by Mr. L. H. Phillips, who collected it from the hills of Bukidnon, 
Mindanao. The flowers are white except the spur and the throat 
of the labellum which are apricot yellow. 

This species is allied to Dendrobium Fairchildae Ames and 
Quis., but differs from it in having smaller and more-slender 
stems, straight spur, and broadly ovate-oblanceolate labellum. 

This species is dedicated to Mr. L. H. Phillips, formerly of the 
Philippine Packing Corporation at Bukidnon, Mindanao. While 
he was there, he spent considerable time and effort in collecting 
orchids during week ends and holidays, and we owe to him our 
present knowledge of the orchid flora of Bukidnon. 

DENDROBIUM SCHUETZEI Rolfc. Plate 1. fiff. 2; Plate 3. figs. 20 to 27; Plate 6, fig. 2. 

Dendrobium Schuetzei Rolfe in Orch. Rev. 19 (1911) 224, 20 (1912) 
289, 308, 337, fig. J^T, in Gard. Chron. Ill 50 (1911) 42, III 52 
(1912) 229, fig. 102; G. Wilson in Orch. World 3 (1912) 19, t; 
Rolfe in Bot. Mag. 139 (1913) t 8^95; Ames Orch. 5 (1915) 
138, 7 (1922) 96, in Merr. Enum. Philip. Fl. PI. 1 (1924) 354. 

The original description reads as follows : 

"Dendrobium Schuetzei Bolfe. — This is another striking Dendrobium, 
of the D. Dearei group, which has been introduced by Messrs. Sander & 
Sons, St. Albans, and of which a technical description has been pre- 
pared. It has very large white flowers, the petals are obovate-orbicular 
in shape, and the lip strongly three-lobed, with broadly-rounded side 
lobes, and the front lobe broadly obovate or nearly orbicular, with a dis- 
tinct apiculus. The petals and lip are much broader than in D. Dearei, 
and altogether different in shape. It should prove a great acquisition." 
Rolfe, Orch. Rev. 19 (1911) 224. 

Rolfe's Latin diagnosis, Bot. Mag. (1913) t. 8495, reads as 
follows : 

Herba epiphytica, 15-40 cm. alta. Caules erecti, subcylindrici, medio in- 
crassati, sulcati, basi attenuati, dense foliati. Falia subpatentia, elliptico- 
oblonga, obtusa, coriacea, 8-10 cm. longa, 2.5-3.5 cm lata. Pedunculi 
subterminales, breves, pauciflori. Bracteae oblongae, subacutae, breves. 
Pedicelli circiter 4 cm longi. Flares magni, speciosi, albi, labelli basi vi- 
ridi. Sepala subpatentia; posticum oblongo-lanceolatum, acuminatum, 3 
cm. longumj lateralia triangularia, acuta, carinata, 3-5 cm. longa; men- 
turn obtusum, 1.3 cm. longum. Petala late ovato-orbicularia, apiculata, 
4.5-5.5 cm. longa, 3.5-4 cm. lata. Labellum trilobum, 4-4.5 cm. longum; 
lobi laterales subincurvi, late rotundati; lobus intermedins subrecurvus, 
late obovatus, truncatus vel emarginatus, apiculatus, crenulatus, 3.5-4 cm. 
latus; discus basi obtuse carinatus. Columna lata, 6 mm. longa; alae fal- 

56,4 Ames and Quisumbing: Philippine Orchids, V 459 

Dwarf in habit; the stems erect, leafy, short, stouter at the 
middle, narrowed to the base, 15 to 30 cm long or more, 1.1 
to 1.5 cm in diameter at the widest portion. Leaves oblong- 
lanceolate or elliptic-lanceolate, subacute or obtuse at the apex, 
narrowed to the base, subcoriaceous, 6 to 11 cm long, 1.6 to 3 cm 
wide. Raceme 3- to 5-flowered ; bracts triangular, about 2 miii 
long. Flowers showy, odorless, 6 to 6.5 cm across, pure white 
except the throat of the lip which is emerald green with a few 
purple stains at the base. Pedicellate ovary greenish white, 5.5 
to 6 cm long. Lateral sepals triangular, acutely acuminate, 
keeled, 3.5 to 4 cm long, 1.5 to 1.8 cm wide. Dorsal sepal lan- 
ceolate, sharply acuminate, 3.5 to 3.8 cm long, about 2 cm wide 
along the column-foot. Petals suborbicular-obovate, broadly 
obtuse or rounded, apiculate, 4 to 4.7 cm long, 3.8 to 4.2 cm 
wide. Labellum distinctly 3-lobed; lateral lobes semiobovate, 
broadly rounded, somewhat incurved, about 2.7 cm long and 1.1 
cm wide near the apex ; middle lobe somewhat recurved, broadly 
flabellate-obovate or nearly orbicular, truncate, or slightly retuse, 
apiculate, 2.4 to 2.6 cm long, 3.2 to 3.7 cm wide; the whole 
labellum pure white except the throat between the lateral lobes 
which has a patch of full green-yellow with short Victoria lake 
lines at the base. Spur white, pale olivine at the tip, short, 
obtuse, 1.3 to 1.4 cm long. Column broad and short, white, 
green dotted with Victoria lake at the base. Anther subquadrate 
in outline, white. PoUinia narrowly oblong. 

Luzon, Manila, Bureau of Science orchid house, Phil Nat 
Herb. 82 Quisumbing, February 13, 1933, originally collected by 
Mr. L. Hachero from Surigao, Mindanao, presented to Doctor 
Quisumbing, and now growing in Bureau of Science orchid house. 
Mindanao, Surigao, W. Lyon s. n., February, 1917. 

A species with the habit of Dendrobium Dearei Reichb. f . and 
D. Sanderae Rolfe and allied to them, but differing strikingly 
in the short, obtuse spur. In addition to the different spur, 
this species differs from D. Dearei in having much larger flowers 
with different petals; it differs from D. Sanderae in the color 
and details of the flowers. 


BULBOPHYLLUM LOBBII Lindl. Plate 2. fiflr. 1; Plate 4, fiffs. 1 to 8; Plate 7» fi^. 1. 
Bulbophyllum Lobbii LiNDL. in Hot. Reg. 33 (1847) sub e. 29 (as 
Bn Lobii) ; W. J. HooKER in Curtis's Bot. Mag. 76 (1850) t. U5S2. 

460 The Philippine Journal of Science ms 

The original description reads as follows: 

BoLBOPHYLLUM LoBii; folio petiolato obovato-oblongo coriaceo, pedun- 
culo unifloro folio breviore, pedunculo nudo unifloro folio breviore, sic basi 
subglanduloso e bracteis squamaeformibus cucullatis falcatis subglandu- 
losis erumpente, sepalis oblongis acutis lateralibus falcatis, petalis con- 
formibus minoribus reflexis, labello longe unguiculato cordato ovato acuto 
canaliculato apice recurvo. Lindl. 

Rhizome creeping, rather stout, 5 to 7 mm in diameter, with 
numerous simple or rarely branched roots which are 1 to 1.5 mm 
in diameter. Pseudobulbs pyriform-cylindric, asymmetrically 
curved, remote, rugose, monophyllous, sessile, 4 to 4.5 cm long, 
1.7 to 1.9 cm in diameter at the widest portion near, the base. 
Leaves oblong-elliptic, acute, coriaceous, rigid, about 12 cm long, 
4.6 cm wide; the petioles about 3 cm long. Scape arising from 
the side of each pseudobulb, its base sheathed with imbricated, 
lanceolate, membranaceous bracts, which are 2.2 to 2.8 cm long; 
the peduncles provided with bracts, which are 1.5 to 2 cm 
long. Pedicellate ovary primuline yellow and minutely spotted 
with dark purplish brown, 4.5 to 5 cm long, the ovary conspic- 
uously ridged. Flowers showy, the largest of all known Phil- 
ippine bulbophyllums, solitary, 7.5 to 8.5 cm across. Lateral 
sepals strongly falcate, lanceolate, gradually narrowed to the 
acute apex, 7.8 to 8 cm long, about 2 to 2.3 cm wide at the widest 
portion near the base, 13-nerved, the anterior portion circinate. 
Dorsal sepal greatly elongated, narrowly lanceolate, gradually 
narrowed to the acute apex, about 9.5 cm long, 1.5 cm wide at 
the widest portion near the base, 11-nerved. Petals similar to 
the lateral sepals, 7.1 to 7.7 cm long, 1,1 to 1.2 cm wide at the 
widest portion near the base, 9-nerved. Labellum triangular- 
lanceolate, strongly recurved, fleshy, mobile, acuminate, cordate 
at the base, 2.8 to 3 cm long, 1.1 to 1.2 cm wide. Column very 
short, stout, free portion 9 to 10 mm high, laterally winged, 
produced into an elongated curved foot about 1.1 cm long, of 
which the terminal portion is free and bears at its apex the 
mobile labellum. Pollinia subglobose. 

Luzon, Benguet Subprovince, Baguio, Mrs. G. H. Fairchild's 
gardens. Bur. Sci. S5568 Quisumbing, March 23, 1932. The 
living plants were originally collected by Mrs. G. H. Fairchild 
near Baguio. 

Bulbophyllum Lobbii is the largest-flowered Bulbophyllum ever 
reported from the Philippines and is closely allied to Bulbophyl- 
lum ccdossum Ridl. 

56,4 Ames and Quisumbing: Philippine Orchids, V 461 


PHALAENOPSIS MARIAE Bnrbidare var. ALBA Ames and Quia, Tar. nor. Plate 2, figs, 
a and 4; Plate 4, flars. 9 to 17; Plate 7, fi;. 2. 

Haec varietas floribus omnino albidis baud striatis a P. Mariae 

In habit and in leaves similar to the species. The leaves, 
shining on the upper surface, dark green, 22.5 to 24 cm long, 
5.5 to 7.7 cm wide. Scapes slender, sparingly branched, few- 
flowered, 10 to 20 cm long. Flowers odorless, 5 to 8 mm dis- 
tant, 3.5 to 4 cm across, without bars and entirely pure white 
except at their apices which are light green-yellow and the inner 
basal portion of the lateral lobes of the labellum which is em- 
pire yellow. Pedicellate ovary white, about 12 mm long. Lat- 
eral sepals spreading, obliquely elliptic-ovate, obtuse, apiculate, 
1.9 to 2 cm long, 1 to 1.1 cm wide. Dorsal sepal narrower, 
erect, broadly lanceolate-elliptic, obtuse, 1.9 to 2 cm long, 8.5 
to 9 mm wide. Petals lanceolate-elliptic, obtuse, 1.7 to 1.8 cm 
long, 8.5 to 9 mm wide. Labellum fleshy, 3-lobed; lateral lobes 
erect in natural position, obliquely subquadrate-oblong, incurved 
towards the column, white with empire yellow stain on the in- 
ner basal portion, 5 to 6 mm long; middle lobe pure white, 
spatulate-obovate, 11 to 12 mm long, 7 to 8 mm wide at the 
widest portion, prominently keeled in the middle longitudinally, 
the keel conspicuously clothed with hairs on the anterior part. 
Column pure white, about 7 mm long. 

Luzon, Manila, Bureau of Science orchid house. Bur. Sci. 
85707 Quisumbing, June 21, 1933. The description is based 
upon a living plant (now growing in the Bureau of Science 
orchid house), which was donated to the junior author by 
Mr. L. H. Phillips, who collected it from the hills of Bukidnon, 

The variety is characterized by having pure white flowers 
with a complete absence of bars on the sepals and petals. 

The plate in Bot. Mag. t. 6964 shows a very similar middle 
lobe of the lip, and the length of the pedicellate ovary is va- 
riable in specimens in the herbarium of the senior author. 
Genus AERIDES Loureiro 

A£RmES LAWRENCIAE Reichb. f. var. FORTICHII Ames and Quia. var. noY. Plate «. 
fiflT. 2. 

Habitu speciei similis, floribus albidis differt. 
A variety resembling the species in habit and general features. 
Leaves 15 to 22.5 cm long, 2.8 to 3.3 cm wide. Inflorescence 

462 The Philippine Journal of Science isas 

curving, about 42 cm long including the peduncles. Flowers 
fragrant, 3.1 to 3.2 cm across. Sepals and petals white faintly 
washed at center with sea-foam yellow, and devoid of purple at 
the tips. Lateral lobes of labellum white, the tips overlapping 
below the middle lobe ; middle lobe rose-pink, margin denticulate. 
Spur goose green. Column white. Pedicellate ovary pale dull 

Luzon, Manila, Bureau of Science orchid house, Phil. Nat 
Herb. 106 Quisumbing, January 13, 1933. According to Mr. 
L. H. Phillips, formerly of the Philippine Packing Corporation, 
Cagayan de Misamis, Mindanao, who furnished the junior au- 
thor with a living plant in flower, the Hon. Manuel Fortich, rep- 
resentative from Bukidnon, collected two plants from his ranch 
in Bukidnon Province, Mindanao. 

As far as we know, this is the first time a white variety of 
Aerides Lawrenciae has been reported. The variety closely re- 
sembles the species in habit and general features, but differs in 
its white flowers. To Representative Fortich, of Bukidnon 
Province, this interesting and newly discovered variety is de- 

AfiRIDES LAWRENCIAE Reichb. f. var. PUNCTATA Ames and Quis. var. nor. Plate 2, 
flffs. 5 and 6. 

Habitu speciei similis, floribus purpureis punctatis differt. 

In habit and general features this variety resembles the spe- 
cies. The flowers are large, fragrant, about 3 cm across. Pe- 
tals and sepals white tipped with true purple. Lateral lobes of 
the labellum white, dotted conspicuously with true purple, the 
dots extending up to the spur; the front lobe of the labellum 
true purple. The tip of the spur is buckthorn brown. 

Luzon, Manila, Bureau of Science orchid house, Phil Nat. 
Herb. 107 Quisumbing, November 12, 1932. The plant was orig- 
inally collected by Mr. L. H. Phillips in the mountains of Caga- 
yan de Misamis, Mindanao. 

This variety is characterized by the conspicuous purple dots 
on the sides of the lateral lobes of the labellum. 
Genus RENANTHERA Loureiro 

RENANTHERA EL0N6ATA Lindl. Plate 2» fiff. 7; Plate 4» figrs. la to 25; Plate 8. 

Renanthera elongata LiNDL. Gen. & Sp. Orch. PL (1833) 218; Ames 
Orch. 5 (1915) 224, Bot. Mus. Leafl. Harv. Univ. 2 (1934) 31. 

Renanthera micrantha Blume, Rumphia 4 (1848) 54, Mus. Bot. Lugd.- 
Bat. 1 (1849) 60; Reichb. p. in Walp. Ann. 3 (1852) 566; MiQ. 
Fl. Ind. Bat. 3 (1859) 698; Naves, Novis. App. (1882) 240. 

56, 4 Ames and Quisumbing: Philippine Orchids, V 463 

Renanthera matutina Lindl. in Bot. Reg. 29 (1843) t 41, G. K. in 
Bot. Zeit. 1 (1843) 760. 
"Caule ramoso, foliis lato-linearibus oblique emarginatis, paniculis elon- 
gatis nutantibus, sepalis exterioribus lateralibus internis latioribus spa- 
tulatis, labelli limbo ovato basi bicalloso. 

"Aerides elongatum. Blume 1. c. [Bijdr. 366.] 

'*Hab. in Java, in rupibus calcareis prope Kuripan, Blume. Flores pu- 

Epiphytic on trees. In habit similar to Renanthera storiei. 
Stems erect, 210 to 215 cm high. Leaves distichous, greenish, 
oblong-elliptic, unequally bilobed at apex, 6.5 to 11.3 cm long, 
2.7 to 4 cm wide, 3 to 4.5 cm distant. Peduncles 11.5 to 11.8 
cm long. Panicles many-flowered (50 to 70 or more) . Flowers 
small, of the same color as Renanthera stoirei var. philippinensis, 
13 to 14.5 mm long, 9 to 10.5 mm wide. Pedicellate ovary 
slender, 9 to 11 mm long. Lateral sepals asymmetrically spa- 
tulate, clavate, rather abruptly dilated above, 6.5 to 7 mm long, 
3.5 to 4 mm wide above the middle, about 1 mm wide across the 
claw. Dorsal sepal oblong-oblanceolate, obtuse, 7 to 7.5 mm 
long, about 3 mm wide at widest portion. Petals very similar 
to the dorsal sepal but smaller, broadly rounded at the apex, 5.5 
to 6 mm long, 2 mm wide. Labellum relatively small, fleshy, 3.5 
to 3.75 mm long, deeply saccate-spurred at base; 3-lobed at the 
apex; lateral lobes transversely subquadrate, broadly truncate, 
about 1 mm high ; the middle lobe strongly recurved, triangular- 
ovate, subacute, about 1.25 mm long, 1 to 1.10 mm wide; spur 
cylindric-conic, about 2 mm long. Column minute, 1.5 mm long. 
Anther broadly ovoid, about 1.5 mm long, 1.25 mm across. Polli- 
nia four. 

Mindanao, Zamboanga Province, on a small island on the east 
coast of Zamboanga, Mrs. Kenneth B. Day s. n., September, 
1932. The plants were epiphytic on trees in mangrove swamps. 

This species is characterized by its very small flowers. 

Genus VANDA Jones 

VANDA MERRILLII Ames and Quis. var. ROTORII Ames and Quis. rar. nor. Plate 2, fir. 
8; Plate 9, Ag 1. 

Haec varieta floribus omnino badiis neque striatis neque ma- 
culatis a Vanda Merrillii differt. 

In habit and flower parts very similar to the species. The 
flowers are essentially the same in size; the sepals and petals 

464 The Philippine Journal of Science 1935 

are ox-blood red within, and chalcedony yellow on the back; 
the lateral lobes of the labellum pure white; the middle lobe 
of the labellum Vandyke red except the base of the auricles 
which are chalcedony yellow; column naphthalene yellow, and 
the pedicellate ovary white. 

Luzon, Manila, Doctor Rotor's gardens, Phil. Nat. Herb. 109 
Quisumbing, February 9, 1934. 

The plant was originally collected by a friend of Doctor Rotor 
from a tree along the road between Baler, Tayabas Province, 
and Cabanatuan, Nueva Ecija Province, Luzon. 

This variety resembles var. immaculata Ames and Quis. in 
the complete absence of bars or maculations on the flowers, but 
differs in the ox-blood red petals and sepals and Vandyke red 
middle lobe of the labellum. 

This variety is dedicated to Dr. A. B. Rotor, a lover of or- 

Genus TRICHOGLOTTIS Reichenbach f. 

TRICHOGLOTTIS GUIBERTII (Linden and Reichb. f.) Ames and Quis. comb. nov. Plate 
2, fiffs. 9 and U; Plate 4, fisrs. 2« to 35; Plate 10. 

Cleisostoma Guiberti Linden and Reichb. f, apud Reichb. f. in Bot. 

Zeit. 20 (1862) 375, Xenia Orch. 2 (1867) 126, t. U2. 
Vanda Guiberti Lindl. apud Linden and Reichb. f. in synon. 

The present species, which is well figured in Xenia Orch. 
1. c, appears to be referable to the genus Trichoglottis as now 
interpreted, while the concept Cleisostoma Bl. is no longer gen- 
erally upheld by orchidologists. 

Moreover, Trichoglottis Guibertii is certainly allied to T. lu- 
zonensis Ames, both vegetatively and florally. On the other 
hand, the name Staurochilus was founded by Ridley on Tricho- 
glottis fasciata Reichb. f . which had previously been referred by 
Bentham, apparently with logic, to Reichenbach's genus Staurop- 

Such species as Trichoglottis Guibertii, T. luzonensis, T. Daw- 
soniana, T. fasciata, etc., certainly differ from the original con- 
ception of Trichoglottis in having large flowers with scarcely 
developed spur, and in having the inflorescences (more or less 
elongate) either loosely racemose or paniculate. 

It seems highly probable that orchidologists will eventually be 
forced to agree with J. J. Smith, in referring to the genu^s 
Trichoglottis all tiiese allied and intergrading species. 

56,4 Ames and Quisumbing: Philippine Orchids, V 465 

"Affine Cleiaastomati ionosmo Lindl. labelli lamina pandurata, pilosula, 
carina postice bicruri a calcaris fundo in laminae fossam excurrente. 

"Vanda Guiberti Lindl. 

"Panicula multiflora. Flores illis Vandae Roxburghii aequales, colore 
fere Anselliae, extus pallide flaveolo-albi. Sepala cuneato-oblonga, obtusa. 
Tepala subaequalia. Omnia flavida annulis rufis. Labellum basi cum co~ 
lumnae basi connatum auriculis rectangulis hinc unidentatis bidentatisve ; 
lamina pandurata, postice latior, apice emarginata, pilosula. Columna bre- 
vis, apice utrinque unifalcis, falcibus velutinis; lamella ovata apiculata 
cochleata sub fovea. 

"Die Berichte tiber die am 24. September in Briissel veranstaltete 
Ausstellung heben als wesentliche Merkwlirdigkeit diese, vom Hm. Direc- 
tor Linden ausgestellte und zunachst Vanda Guiberti genannte Orchidee 

"In der That ist dieselbe eine sehr merkwiirdige acquisition, besonders 
wenn man erwagt, wie die Cleisostomen bis jetzt keine irgend Mbschen 
Bliithen aufwiesen, es sei denn das doch gar bescheidene CL ionosmum, 
Unsere neue Art hat einen machtigen Bliithenstand von Bliithen, die einen 
modernen deutschen Vereinsthaler decken, und deren braune Einge auf 
Paille Untergrund einen guten Eindruck machen." Reichb. f. 

Coarse, stout, epiphyte, approaching Trichoglottis ionosma in 
habit. Roots stout, elongate, glabrous. Stems elongate, 50 to 
70 cm long, 10 to 14 mm in diameter, subterete. Leaves dis- 
tichous, oblong-ligulate, 14 to 28.5 cm long, 2.5 to 3.3 cm wide, 
coriaceous, 3 to 3.5 cm distant, dark green, unequally bilobed 
at the apex with obtuse lobes. Peduncles rigid, erect when oc- 
curring on erect stems and rather pendent when on drooping 
stems, 4 to 6 mm in diameter. Inflorescences showy, laxly pa- 
niculate, many-flowered, inclusive of the peduncles, 16 to 35 cm 
long, glabrous. Ovary twisted. Flowers fleshy, 3.5 to 4.5 cm 
across. Lateral sepals elliptic-obovate, obtuse, 1.6 to 1.8 cm long, 
9 to 11 cm wide, somewhat asymmetric. Dorsal sepal, obovate- 
oblong, rounded at the apex, 1.6 to 2.1 cm long, 7 to 9 mm wide. 
Petals spatulate, more or less falcate, rounded at the apex, 1.5 
to 1.8 cm long, 7.5 to 9 mm wide above. Labellum trilobed, 
fleshy, shortly saccate-spurred at the base, 1.6 to 2 cm long, 
densely pubescent on both surfaces especially on the inner sur- 
face; lateral lobes erect and connate with the column, short, 
rounded-triangular, 1.5 to 2 mm high; middle lobe much larger 
than the lateral lobes, subquadrate-pandurate, recurved at the 
apex, retuse at the tip, 1 to 1.3 cm long, 8 to 9 mm wide at the 
broadest portion near the base, with a prominent longitudinal 
keel on the upper surface; spur shortly conical, obtuse, 4 to 5 

466 The Philippine Journal of Science 

mm long. Within the saccate base of the labellum (between the 
lateral lobes) there is a small, ovate, abruptly acuminate, deeply 
concave ligule, about 5 mm long. Column short, stout, 7 to 8 
mm long, densely and minutely pubescent, on each side at the 
apex with upcurved falcate stelidia which are 3 to 3.5 mm long. 
PoUinia 4, united in pairs, unequal. 

Luzon, Manila, Mrs. Remedios C. Gonzales's gardens. Bur. Sci. 
8^716 Quimmbing, June 22, 1932. 

The living plants were originally collected on limestone rocks 
and on branches of trees in Quezon National Park, Tayabas 
Province, Luzon. 

The flowers are fragrant, and remain open for about eight 
weeks. Sepals and petals marguerite yellow; barred and spotted 
with Hay's russet or madder brown, pale lumiere green at the 
tips ; lateral lobes of the labellum white outside, primuline yellow 
inside spotted with vinaceous rufous, middle lobe barium yellow 
spotted with vinaceous-ruf ous ; ligule between the lateral lobes 
at the base white; keel of labellum and column white; spur mar- 
guerite yellow ; arms of the column pompeian red ; anthers am- 
ber yellow. 

This species is now reported from the Philippines for the 
first time, ft differs from T. ionosma and T. luzonensis in the 
form of the lip. 


Plate 1 

Pig, 1. Dendrobium anosmum Lindl. var. Dearei (Rolfe) Ames and Quis. 
comb, nov., front view of flower, x 1. 

2. Dendrobium Schuetzei Rolfe, front view of flower, X 1 (circa). 

3. Dendrobium anosmum Lindl. var. Huttonii (Reichb. f.) Ames and 

Quis. comb, nov., X 1. 

4. Dendrobium cerinum Reichb. f., front view of flower, x 1. 

5. Dendrobium cerinum Reichb. f., side view of flower, x 1. 

6. Dendrobium Phillipsii Ames and Quis. sp. nov., front view of 

flower, X 1 (circa). 

7. Dendrobium PhilUpm Ames and Quis. sp. nov., side view of 

flower, X 1 (circa). 

Plate 2 

Pig. 1. Bulbophyllum Lobbii Lindl., front view of flower, X 1 (circa). 

2. Aerides Lawrenciae Reichb. f. var. Fortichii Ames and Quis. var. 

nov., front view of flower, x 1 (circa). 

3. Phalaenopsis Mariae Burbidge var. alba Ames and Quis. var. nov., 

front view of flower, X 1. 

4. Phalaenopsis Mariae Burbidge var. alba Ames and Quis. var. nov., 

side view of flower, x 1. 

5. Aerides Lawrenciae Reichb. f, var. punctata Ames and Quis. var. 

nov., front view of flower, x 1 (circa). 

6. Aerides Lawrenciae Reichb. f. var. punctata Ames and Quis. var. 

nov., side view of flower, X 1 (circa). 

7. Renanthera elongata Lindl., front view of flower, X 4 (circa). 

8. Vanda Merrillii Ames var. Rotorii Ames and Quis. var. nov., front 

view of flower, x 1. 

9. Trichoglottis Guibertii (Linden and Reichb. f.) Ames and Quis. 

comb, nov., front view of flower, X 1. 
10. Trichoglottis Guibertii (Linden and Reichb. f.) Ames and Quis. 
comb, nov., side view of flower, X 1. 

Plate 3 

Dendrobium cerinum Reichb. f.: i, dorsal sepal, X 1.5; 2, petal, X 1.5; 
S, lateral sepal, side view of column, pedicellate ovary, and spur, 
X 1.5; 4, labellum from above (expanded), X 1.5; 5, side view 
of labellum (natural position), X 1.5; 6, front view of column, 
X 1.5; 7, anther from above, X 3; 5, anther from below, x 3; 
9-10, pollinia, X 6. 


468 The Philippine Journal of Science i93s 

DendroMum Phillipsii Ames and Quis. sp. nov.: 11, dorsal sepal, X 1 (cir- 
ca) ; 1^, petal, X 1 (circa); IS, lateral sepal, X 1 (circa); 
i4, labellum from above (expanded), X 1 (circa) ; 15, side view 
of ovary and column, X 2.5 (circa) ; 16, front view of column, 
X 2.5 (circa) ; IT, anther from above, X 4.5 (circa) ; 18, an- 
ther from below, x 4.5 (circa) ; 19, poUinia much enlarged. 

DendroMum Schuetzei Kolfe: SO, dorsal sepal, X 0.5 (circa) ; 21, lateral 
sepal, X 0.5 (circa) ; 22, petal, X 0.5 (circa) ; 2S, side view of 
ovary, column, and labellum (natural position), X 1 (circa); 
24, labellum from above (expanded), X 0.5 (circa); 25, front 
view of colunm, X 1 (circa); 26, anther from above, X 2.5 
(circa); 27, anther from below, X 2.5 (circa). 

Plate 4 

Bulbophyllum Lobbii Lindl.: 1, dorsal sepal, X 0.5 (circa); 2, petal, X 
0.5 (circa) ; S, lateral sepal, X 0.5 (circa) ; J^, front view of 
column and labellum (natural position), X 0.5 (circa); 5, side 
view of ovary, colunm, and labellum (natural position), X 0.5 
(circa); 6, labellum from above (natural position), X 2 (cir- 
ca); T, anther from above, very much enlarged; 8, anther from 
below, very much enlarged. 

Phalaenopsis Mariae Burbidge var. alba Ames and Quis. var. nov.: $, dorsal 
sepal, X 1 (circa) ; 10, petal, X 1 (circa) ; 11, lateral sepal, 
X 1 (circa); 12, side view of column and labellum (natural 
position), X 2 (circa); 13, labellum from above (expanded), 
X 2 (circa) ; i4, front view of column and labellum (natural 
position), X 2 (circa); 15, anther from above, X 6 (circa); 
16, anther from below, X 6 (circa); 17, poUinia, X 6 (circa). 

Benanthera elongata Lindl.: 18, dorsal sepal, x 2.5 (circa); 19, petal, 
X 2.5 (circa) ; 20, lateral sepal, X 2.5 (circa) ; 21, column and 
labellum from above, (natural position) , x 3.5 (circa) ; 22, side 
view of ovary, column, and labellum (natural position), X 3.5 
(circa) ; 2S, anther from above, x 10 (circa) ; 24, anther from 
below, X 10 (circa) ; 25, pollinia, X 10 (circa) . 

Triehoglottis Guibertii (Linden and Reichb. f.) Ames and Quis. comb, nov.: 
26, dorsal sepal, X 3 (circa) ; 27, lateral sepal, X 3 (circa) ; 
28, petal, X 3 (circa) ; 29, side view of column and labellum 
(natural position), x 3 (circa); SO, front view of column, x 
3 (circa) ; SI, column and labellum from above (natural posi- 
tion) X 3 (circa); S2, labellum from above (expanded), X 
3 (circa); SS, anther from above, very much enlarged; S4, an- 
ther from below, very much enlarged; S5, pollinia, very much 

FhATB 5 

Dendrobium anasmum Lindl. var. Dearei (Rolfe) Ames and Quis. comb, 
nov., habit, very much reduced. 

Plate 6 
Fia. 1. Dendrobium Phillipsii Ames and Quis. sp. nov., flowers, approx- 
imately natural size. 
2. Dendrobium Schuetzei Rolfe, habit, very much reduced. 

56,4 Ames and Quisumbing: Philippine Orchids, V 469 

Plate 7 

Fig. 1. Bulhophyllum Lobbii Lindl., habit, much reduced. 

2. Phalaenopsis Mariae Burbidge var. alba Ames and Quis. var. nov., 
habit, very much reduced. 

Plate 8 
Renanthera elongata Lindl., habit of the plant, very much reduced. 

Plate 9 

Fig. 1. Vanda Merrillii Ames and Quis. var. Rotorii Ames and Quis. 
var. nov., habit, very much reduced. 
2. Dendrobium cerinum Reichb. f., habit, very much reduced. 

Plate 10 

Trichoglottis Guibertii (Linden and Reichb. f.) Ames and Quis. comb, 
nov., habit, very much reduced. 

291458 5 

Amks am> Qi^; ; Piiii.ii'fiNi-; Owrmus. VJ 

; I'lfii.ii*, ,h)i,:ti!v. Sii., HH, M'S 


AMKH ANfi t|l-fSf'\l|if .\(; ; f'lntJri'IM.; 


Amrs anp QursuMMNG: Philippine Oechiiis. ¥»] 

[Fmmw, JoiJiK. Sor., 56. No. 4. 


Amks anji QrisrwitiNU: I'iiiliwink Oiu'hii»s, V.J 


Ames ani> Qisksumhinu : Piuufpink OHrmii,-.. V. 

I l*iiiUF. J*H KN. Srr., M, Nti. I. 









AhWiiA AKn QvmvmiiNe,: Philippi.n'k Oecoidh, V.| 

[Piittip. .JmniN, BcL, 56, No. i 




<"'S? .^T5 



AMKS AKii C|i:iSUMBI.N«;: PHIUI'PINK Oli<*lll!>.S, V.J 

I |-*»iL!P. .hiTKN. Scu, .'ii;, Nw. 4. 


3 ' i^ 

^-^^ .. .. 




PLATE 10. 



Of the Department of AgricuUure arid Commerce, Manila 

Christensen and Holttum have just published ^ a most valuable 
paper on the ferns of Mount Kinabalu, listing earlier collec- 
tions, those by Mr. Holttum and collectors sent by him, and 
those of Mr. and Mrs. Clemens up to August, 1933. The Cle- 
mens expedition closed its work there in January, 1934. At Mrs. 
Clemens's request, Mr. Holttum being in England and identifica- 
tions being urgently needed to make possible the distribution 
of the collections, I have undertaken to name the ferns collected 
during their last four months in the field. The number of addi- 
tions made in this period is sufficient evidence that this mountain 
is by no means yet fully explored. These additions, with the 
Christensen and Holttum list, raise the number of Kinabalu 
ferns to 437 species. 

The new and locally new species, and a few others inviting 
comment, follow. 


Above Kinitaki River, on wet hill, altitude 4,500 feet, Cle- 
mens SOJflO. Apparently new to Borneo. The same large form 
is in Java and Mindanao. A smaller one in Mindanao and the 
central Philippines is sometimes distinguished as C. cumingiana. 

CYATHEA SUBBIPINNATA Copelaxid sp. hot. Plate 1. 

Trunco, ut videtur, 3 ad 4 cm crasso, verosimiliter humile^ 
apice stipiti conf orme paleis immerso ; stipite 25 ad 30 cm longo,. 
paleis confertis castaneis pallide marginatis minute atrocasta- 
neo-ciliatis deorsum 12 mm longis, 3 mm latis apiculatis utroque 
latere stipitis expansis insigniter ornato; fronde 60 cm longa, 
20 cm lata subbipinnata, rhachi deorsum paleis decrescentibus 
ornata, sursum setulosa; pinnis infimis deflexis, brevi-stipitulatis, 
10 cm longis, medio 23 mm latis, basi pinnata pinnulis saepius 

^The ferns of Mount Kinabalu, Gardens' Bull. 7 (1934) 191-234, pis. 



472 The Philippine Journal of Science 1935 

1 libera et 2 adnatis, alibi pinnatifidis ; pinnis medialibus sessili- 
bus, 10 cm longis, 18 mm latis, acutis, costis glabrescentibus, 
segmentis f alcato-rotundatis, 5 mm latis, nudis ; venulis 6- vel 7- 
paribus, una basiscopica subapicale plerumque furcata caeteris 
simplicibus; soris inframedialibus, castaneis, nudis, 

Borneo, Mount Kinabalu, Gurulau spur, "crest of ridge, alti- 
tude 6,000 feet," Clemens 50396. 

Similar to C. bipinnatifida Copel. of Basilan, which has very 
hairy axes and pale basal scales. The group is that of C. squam^ 
ulata, which has narrower paleae and bullate squamules. This 
is perhaps the bipinnatifid plant referred to by Christensen and 
Holttum at the bottom of page 218, but I am unable to construe 
the plant in hand as a small C. sqvmrndata or C. eUiptica. 

I mistrust, though, that we have a better developed specimen 
of this species in Clemens 40354 from Dahobang River ridge, al- 
titude 3,500 feet. It is a tree-fern 10 feet tall with tripinnatifid 
fronds on long stipes, the apical part, of course, bipinnatifid. 
The lobes are narrower than in C snhhipinnata as just described, 
more falcate, with three or four forked veinlets on the convex 
side, rather acute, the venation closer, and the veinlets salient on 
the upper surface. These differences might be correlated with 
greater exposure. The paleae are somewhat paler, and the 
double row of them extends up the rachis almost as far as there 
are pinnae. It has the aspect of C. sqitamulata, but is sharply 
enough distinguished from this by its nakedness. 

CYATHEA HOLTTUMII Copeland sp. nov. Plate 2. 

C. squamulatae affinis, trunco 1.5 m alto; stipite 45 cm longo, 
infeme paleis fuscis lanceolatis 5 mm longis rigide appressis 
vestito, lateribus paleis lanceolatis apice protractis nitidis fuscis 
patentibus integris densissime onusto; fronde 1.8 m (teste lec- 
tore) longa, ovata, apice excepta tripinnatifida, rhachi fusca, in- 
ferne setis ferrugineis plerisque 5 mm rarius usque ad 1 cm 
longis vestita, alibi densissime more stipitis paleacea; pinnis 
infimis 25 cm longis, 8 cm latis, rhachilla infeme setosa ; media- 
libus 45 cm longis, 13 cm latis, rhachilla inferne glabrata, sessi- 
libus, acutis ; pinnulis subsessilibus basi truncatis, 6.5 cm longis 
14 mm latis, acutis, | ad costas pinnatifidis, papyraceis, inferne 
pallidis, costis inferne paleis albidis 0.5 ad 1 mm longis ovatis 
et lanceolatis adspersis, lobis 3 mm latis, approximatis, oblique 
subacutis, costulis squamulis paucis deciduis albis vestitis ; ve- 

56, 4 Copeland: Additional Ferns of Kinabalu 473 

nulls ca. 5-paribus quarum 1 vel 2 infra marginem convexum 
furcatis; soris fere medialibus, nudis. 

Borneo, Mount Kinabalu, "ridge south of Dahobang River, 
altitude 3,500 feet," Clemens U0308. 

It is noted that this and the larger plant ascribed to C sub- 
bipinnxita are from the same locality. This one is distinguished 
by nonciliate paleae, the many long setae on the nether face of 
the rachis, and squamules on costae and costules. These dis- 
tinguish it from C. squamulata. I do not know Alsophila par- 
vifolia Holttum, which this may be; at any rate, that specific 
name is not valid for it in Cyathea. 


Species aspectu Hypolepidis punctatae soris dorsalibus stipite 
paleaceo: rhizomate 1 cm crasso, adscendente, basibusque stipi- 
tum paleis rufis 2 ad 4 mm longis lanceolatis apice attenuatis 
sat dense vestitis; stipitibus approximatis non caespitosis, 45 
ad 60 cm altis 2 ad 3 mm crassis, stramineis, paleis sparsis mi- 
nutis aut lanceolatis aut in fibrilla angustatis vestitis; fronde 
anguste ovata, 60 ad 70 cm alta, tripinnata, membranacea, rha- 
chibus m